Implementing QoS for VoIP in a Local Area Network (LAN)
Contents
1. N Wireless Workgroup Multilayer Router Access Point W switch Switch J Figure 5 2 Logical LAN Network Topology It is necessary to understand that there are three regions in this LAN topology Region 1 region 2 and region 3 The LAN in region 1 is extended to region 2 and 3 using an IP MPLS based VPN from a service provider The network diagram above is made up of different devices with different IOS models and software versions and knowing this is important because implementing QoS in a network will be a simple task if the internetworking devices offer the right QoS features and it will be difficult if it does not TURKU UNIVERSITY OF APPLIED SCIENCES BACHELOR S THESIS Donald Egbenyon 30 Table 5 1 Devices in the network configured Equipment IOS Model IOS Software Version Cisco Catalyst 2950 switch WS C2950G 24 EI 12 1 22 EA4a WS C2950G 48 EI 12 1 22 EA6 Cisco Catalyst 2960 switch WS C2960G 24TC L 12 2 25 SEE2 WS C2960G 48TC L 12 2 35 SE5 WS C2960G 8TC L 12 2 44 SE6 Cisco Catalyst 3524 switch WS C3524XL 12 0 5 WC17 Cisco Catalyst 4500 series Sup 11 TS WS 12 2 53 SG2 X4013 TS SNOM Hard phone SNOM 320 VIP Softphone 3 4 The switch in region 1 is made up of C2950 C2960 C3524 and C4500 while those in region 2 are made up of C2960 only Region 3 is made up of both C2960 and C2950 Since there are different Cisco switches in t2. Extended ping 1 100 100 Extended ping 2 100 100 Extended ping 3 100 100 Table 6 4 192 168 248 51 to 193 142 250 8 QoS Baseline Ping results for Region 1 to Region 2 Packet Size WithOut QoS Policy With QoS Policy 160 bytes min avg max ms min avg max ms Extended ping 1 8 15 25 1 7 25 Extended ping 2 8 15 25 1 7 17 Extended ping 3 8 15 42 1 8 17 Success rate Success rate Extended ping 1 100 100 Extended ping 2 100 100 Extended ping 3 100 100 TURKU UNIVERSITY OF APPLIED SCIENCES BACHELOR S THESIS Donald Egbenyon 38 While the results for the Pings to the switches in Region 1 may not clearly show the effect of the QoS configuration on a LAN the result of the ping to the switches in the region 2 shows that because the Ping was moved to the expedited forwarding class with the VOIP traffic the minimum average and the maximum response time were shorter before the QoS setting was implemented So it can be clearly seen that the VoIP traffic in this network is being given expedited forwarding therefore QoS configuration is beneficial to this network traffic from region 1 to region 2 Below is an output from one of the switches to show the packets that matched the different access lists and are being given expedited forwarding to further buttress the fact that the voice packets are given expedited forwarding It can be seen from the output below that
3. Keywords VoIP QoS LAN DiffServ CONTENTS 1 Introduction 2 QoS Overview 2 1 Models of QoS 3 QoS for Voice 3 1 00S Tools for VoIP 3 1 1 Classification 3 1 2 Queuing 3 1 3 Network Provisioning 4 Gathering Information 4 1 Deductions from analysed data 5 QoS Implementation 5 1 QoS Model used 5 1 1 Preparing to implement a QoS model 5 1 2 Implementing the chosen QoS model 5 2 Lab test 5 3 Configuration used 5 4 Difficulties caused by Cisco IOS 6 Results and Conclusion 6 1 QoS Implementation 6 1 1 QoS baseline 6 1 2 QoS baseline results 6 2 Conclusion 14 14 15 22 23 23 25 25 26 28 31 34 34 34 35 40 7 References 8 Appendix FIGURES Figure 2 1 Overview of the TAG and User Priority ina Layer 2 802 1Q frame 26 Figure 3 1 Figure 3 2 Figure 3 3 Figure 3 4 Figure 3 5 Figure 3 6 Figure 4 1 Figure 4 2 Figure 4 3 Figure 4 4 Figure 5 1 Figure 5 2 Figure 6 1 Figure 6 2 TABLES IPv4 and IPv6 headers 5 The original IPv4 ToS Byte 5 DiffServ CodePoint Field 5 SNOM QoS setting 19 TOS 8 bits field 18 Comparing ToS to IP Precedence and DSCP 18 Captured Phone call Traffic to Handset TCP Communication Analysing TLSv1 protocol Using UDP for VoIP communication Lab Logical Topology Logical LAN Network Topology Region 2 switch Policy for an interface Table 2 1 Comparing IP Precedence and DSCP values Table 3 1 Comparing IP Pr
4. CRITIC ECP 0100 maximize throughput 100 Flash Override 0010 maximize reliability 011 Flash 0001 minimize monetary cost 101 Immediate 001 Priority 000 Routine TURKU UNIVERSITY OF APPLIED SCIENCES BACHELOR S THESIS Donald Egbenyon 11 Figure 3 2 The original IPv4 ToS Byte 5 This figure shows the DS field as is presently used today in the IPv4 header Bits 0 1 DS Field Class Selector i Codepoints i Currently i Unused Differentiated Services Codepoint DSCP RFC 2474 Figure 3 3 DiffServ CodePoint Field 5 There is something worthy of note when it comes to classification Some companies use the TOS marking to mark their DSCP settings For example most SNOM phones are marked with a TOS setting of 160 This can be seen in the figure below Quay Of Service Satins Sent ATP Type savi Quilty of Service ATP Type of Seve TOSDiKen 10 i SP Tyget Sane SIP Tye Serve TOSO 10 i Figure 3 4 SNOM Oos setting 19 TURKU UNIVERSITY OF APPLIED SCIENCES BACHELOR S THESIS Donald Egbenyon 12 Initially this looks confusing because it is not clearly started that this is IP precedence 5 until a close look is taken into ToS Type of Service ToS is an 8 bit field in the IP datagram header It has been in the IP header from the beginning but was not used It has two parts the IP precedence value and ToS bits Figure 3 5 below gives a clear picture of the location of thi
5. Inc 2005 Configuring quality of service Catalyst 4500 series switch cisco IOS software configuration guide 12 2 25 SG ed pp 406 458 San Jose California Cisco Systems Inc 5 Cisco Systems Inc 2005 DiffServ the scalable end to end QoS model No 19 San Jose CA Cisco Systems Inc 6 Cisco Systems Inc 2006 Configuring QoS Catalyst 3750 switch software configuration guide 12 2 25 SEE ed pp 703 781 San Jose CA Cisco Systems Inc 7 Cisco Systems Inc 2008 Configuring QoS Catalyst 2950 and catalyst 2955 switch software configuration guide 12 1 22 EA11 ed pp 529 567 San Jose CA Cisco Systems Inc 8 Cisco Systems Inc 2010 Legacy QoS CLI commands deprecation Retrieved August 10 2011 from http www cisco com en US prod collateral iosswrel ps6537 ps6558 product bu lletin c25 580832 html TURKU UNIVERSITY OF APPLIED SCIENCES BACHELOR S THESIS Donald Egbenyon 42 9 Flannagan M E Durand B Sommerville J Buchmann M amp Fuller R 2001 Administering cisco QoS in IP networks Rockland MA Syngress Publishing Inc 10 Gallon C 2003 Quality of service for next generation voice over IP networks No 17 California Multiservice Switching Forum 11 Gonia K 2004 Latency and QoS for voice over IP No 21 SANs Institute 12 JHT2 2011 H 323 Retrieved July 27 2011 from http www voip info org wiki view H 323 13 Park P 2009 Voice
6. Per Hop Behaviour Access Control Entity 1 Introduction A local area network LAN is a network covering a small geographical area LAN is a high speed data network that uses the Ethernet technology in connecting different devices together Since the Ethernet technology offers simplicity it is being deployed in WAN and used to connect LAN thus bringing simplicity to the Internet as well The Internet is a system of interconnected networks all over the world that connects millions of computers together The Internet network was designed to carry data on a best effort delivery basis using either Transmission Control Protocol TCP or User Datagram Protocol UDP which are the two most common transport layer protocols of the Internet It was designed with the notion of giving all traffic equal priority over the network All packets have an equal chance to be delivered and an equal chance also to be dropped This best effort model of the Internet Protocol IP was completely suitable for most data packets carried over the network from one point to another until the advent of Voice over Internet Protocol VoIP or IP Telephony Thus the network should be able to carry voice data video and other traffic over a converged network instead of creating and maintaining separate networks A converged network allows the flow of the voice video and data packets over it This convergence reduces cost and complex network layers When it is carefully des
7. If this is not done it will eventually lead to chaos in the network as a result of congestion This congestion is as a result of too many packets competing for the same bandwidth thus all packets have an equal chance of being dropped by the switch or router To avoid such scenario Quality of Service QoS has to be configured on the switch in order for preferential treatment to be given to packets that are extremely important and require less delay In this thesis priority is given to voice packets in a particular network by configuring the switches to recognize voice packets The goal is to implement QoS on the entire network The purpose of this thesis is to configure QoS in a converged local area network It starts with giving an overview of QoS in Chapter 2 This chapter explains the reason QoS is needed for a network Chapter 3 then gives details of how QoS can be beneficial to VolP while Chapter 4 is used to explain the necessary materials required to configured QoS in a specific network Chapter 5 then goes further to describe the way QoS configuration can be implemented in a network while Chapter 6 describes the results gathered and the conclusion of the work TURKU UNIVERSITY OF APPLIED SCIENCES BACHELOR S THESIS Donald Egbenyon 2 QoS Overview Quality of Service QoS is the ability of a networking equipment to differentiate among different classes of traffic and to give each class different priority over the network when ther
8. devices and the providers devices All that the provider devices do is to carry Layer 2 frames from one area like region 1 to another like region 2 in a manner transparent to the customer edge devices So QoS should have been implemented in the PE router by the service provider 5 4 Difficulties caused by Cisco IOS Something of note is the various difficulties experienced as a result of the Cisco IOS limitation Even with an upgraded software version there are still so many restrictions in some of the switches used in this network In C2950 switches there is a number of restrictions 2 e All ACE in an ACL must have the same user defined mask Additionally on a given interface only one type of user defined mask is allowed e Only four user defined masks can be defined for the entire system It can be used by either security or quality of service QoS but cannot be shared by QoS and Security This is the most inconvenient restriction experienced when implementing the QoS configurations on these switches TURKU UNIVERSITY OF APPLIED SCIENCES BACHELOR S THESIS Donald Egbenyon 32 Table 5 2 Summary of ACL Restrictions Restriction Number of U Number of A Number Permitted ser defined masks allowed in an ACL 1 CLs allowed on an interface 1 and QoS allowed on a switch Total number of user defined masks for security 4 In C2960 switches and below there is a limitation on the use of show polic
9. for IP packets but it requires special QoS to be made available by the network for a period and that bandwidth should be reserved This method uses a protocol to reserve bandwidth on a per flow basis This protocol is called the Resource Reservation Protocol RSVP RSVP is a signalling mechanism that is used by IntServ architecture to carry out its function Once the IntServ session is established it has to be maintained by the router along the path to that session The IETF recommends that RSVP path and Reservation messages should be sent every 30 seconds periodically along the session path to prevent the soft state from timing out in the routers A session will continue until it is torn down or there is no refresh messages received by the routers along the path and when this happens the soft state TURKU UNIVERSITY OF APPLIED SCIENCES BACHELOR S THESIS Donald Egbenyon in the router times out When such a measure is implemented packet delivery is guaranteed but it limits the scalability of a network On the contrary DiffServ provides the need for simple and coarse methods of putting traffic into classes called Class of Service COS It does not specify that a specific protocol should be used for providing QoS but specifies an architectural framework for carrying out its function DiffServ carries out its major function through a small well defined set of building blocks from which different aggregates of behaviours can be built 5 The
10. over IP security Indianapolis Ind Cisco Press 14 Persky D 2007 VoIP security vulnerabilities No 127 SANS Institute 15 Provisioning 2000 Retrieved July 19 2011 from http searchsoa techtarget com definition provisioning 16 Session initiation protocol 2010 Retrieved July 20 2011 from http en wikipedia org wiki Session Initiation Protocol 17 Snom Technology A 2005 Snom 320 user manual 1 00th ed Berlin Germany 17 Snom Technology A 18 Snom Technology A 2009 FAQ What does the TOS values 160 mean Retrieved July 27 2011 from http wiki snom com FAQ What_ does the TOS value 160 mean 19 Snom Technology A 2009 Web Interface V8 Advanced Retrieved July 14 2011 from http wiki snom com Web_Interface V8 Advanced 20 Transport layer protocol 2010 Retrieved July 20 2011 from http en wikipedia org wiki Transport Layer Security 21 Tucker G S 2004 Voice over internet protocol VoIP and security No 16 SANS Institute TURKU UNIVERSITY OF APPLIED SCIENCES BACHELOR S THESIS Donald Egbenyon 43 22 Wallace K CCNP 2005 Voice over IP first step Indianapolis Ind Cisco 23 Wallace K CCNP 2009 Authorized self study guide Cisco voice over IP CVOICE 3rd ed Indianapolis IN Cisco Press 24 Wallace K CCNP 2011 In Wallace K CCNP Authorized self study guide Ed Implementing cisco unified communications voice over IP and Qo
11. serve in different roles like multimedia communication voice video and data conferencing and inter working with PSTN applications 11 It is the most widely used VoIP signalling and call control protocol today It has four types of elements defined in this protocol terminal gateways gatekeepers and Multipoint Control Units MCU Every end user device running the H 323 protocol is a terminal and gateways connect the H 323 network to other networks like PSTN A gatekeepers provide services like addressing authorization and authentication for each terminal and gateway It also manages the bandwidth utilization by all devices The MCU allows multiparty audio or video conferences between several H 323 terminals d SIP Session Initiation Protocol SIP is an IETF defined protocol for VoIP text and other multimedia sessions 16 It is used for creating modifying and terminating sessions between one or more participants The major difference between SIP and other signalling protocols is that it is a text based application layer protocol and simpler than H 323 It does not require the use of Call agent when commencing and concluding between two clients Since it works like HTTP in carrying out communication between clients it became popular for companies that do not need complex VoIP setup Since it is an application protocol it can be carried out by UDP TCP and STCP protocols Its simplicity might have made companies like Sonera to choo
12. to be used is DiffServ it is necessary for the network engineer to decide how this model will be implemented This model can be implemented using Cisco AutoQoS or the Modular QoS CLI MQC Cisco AutoQoS is a Cisco proprietary of implementing QoS on Cisco devices It can only be used on Cisco router s or switches but it is not supported by all switches AutoQoS makes assumptions about the network design and generates a set of configurations suitable for the devices in the network It is very easy to deploy and generates an efficient configuration suitable for the device and network The major disadvantage of using AutoQoS in a heterogeneous network is that it recognises only Cisco devices TURKU UNIVERSITY OF APPLIED SCIENCES BACHELOR S THESIS Donald Egbenyon 26 Modular QoS CLI MQC is the most efficient way to implement QoS on any networking device like Cisco HP etc in a heterogeneous network It is a common set of configuration commands used to configure most QoS features in a router or switch The modularity is excellent and network administrators have complete control over the configuration to be used in the router or switch Since QoS is being implemented in a heterogeneous LAN it will suffice if some of the steps are left to default and Modular QoS CLI is used to implement DiffServ in the network The fact is that AutoQoS will not be able to recognise the range of ports to be configured and it will not be able to recognise the
13. 0 40 AF12 Medium 001100 0x0C 12 48 AF13 High 001110 Ox0E 14 56 2 010 CS 2 010000 0x10 16 64 AF21 Low 010010 0x12 18 72 AF22 Medium 010100 0x14 20 80 AF23 High 010110 0x16 22 88 3 011 CS 3 011000 0x18 24 96 AF31 Low 011010 Ox1A 26 104 AF32 Medium 011100 0x1C 28 112 AF33 High 011110 Ox1E 30 120 4 100 CS 4 100000 0x20 32 128 AF41 Low 100010 0x22 34 136 AF42 Medium 100100 0x24 36 144 AF43 High 100110 0x26 38 152 5 101 CS5 101000 0x28 40 160 EF 101110 Ox2E 46 184 TURKU UNIVERSITY OF APPLIED SCIENCES BACHELOR S THESIS Donald Egbenyon 14 Presently IP Precedence is still used by IP devices to mark packets This is a transitional phase because most IP devices do not yet support DSCP The ideal situation is that IP devices will use the DSCP value of Expedited Forwarding EF for RTP voice and VOIP Control traffic will have a DSCP value of Assured Forwarding 31 AF31 in the future 3 1 2 Queuing Every packet that will be forwarded will have to be placed in a queue Based on classification queuing tools assign packets to several queues for the required treatment in the network The voice data and video are placed in different queues on egress interfaces based on the classification This will ensure that priority is given to the needed packets that need priority This will be explained more with class based queuing priority queuing when QoS is implemented 3 1 3 Network Provisioning Network provisi
14. 48 121 E Ci 45 6 413645 62 71 106 251 192 168 248 121 60 sip tls gt 53381 ACK Seq 3435 Ack 1093 46 6 413675 62 71 106 251 192 168 248 121 TLSV1 128 Application Data Application Data 47 6 433201 192 168 248 121 192 168 248 55 ICMP 74 Echo ping request id 0x0001 seq 150 48 6 439048 192 168 248 55 192 168 248 121 ICMP 74 Echo ping reply 1d 0x0001 seg 150 49 6 501411 vmware ac 00 0d Broadcast 60 who has 192 168 248 397 Tell 192 168 2 52 6 974179 HonHaiPr 9b 3c f0 60 who has 192 168 248 587 Tell 192 168 2 Y gt stream index 5 n Sequence number 1 relative sequence number L Acknowledgement number 2761 relative ack number Header length 20 bytes 8 Flags 0x10 ACK window size value 258 calculated window size 258 window size scaling factor 1 Cunknown Checksum 0x627f validation disabled 8 SEQ ACK analysis This is an ACK to the segment in frame 35 i The RTT to ACK the segment was 0 000028000 seconds 0000 00 14 a9 c3 55 ff fo de f1 Of c7 6d 08 00 45 00 0010 00 28 Oe 3a 40 00 80 06 00 00 co a8 f8 79 3e 47 0020 6a fb do 85 13 c5 bl 64 25 10 1c 52 f8 85 50 10 0030 01 02 62 7f 00 00 Broadcast Fie C Documents and Settings Ehis My Docum Packets 1249 Displayed 1249 Marked O Load time 0 00 500 Profile Default Figure 4 2 TCP Communication From Figure 4 2 analysing frame number 36 shows that after a session has been initiated by the VIP sof
15. Address 6 IP Address 7 IP Address 8 IP Address 9 IP Address 10 IP Address Tallenna ToS bit used in SNOM O p lle O pois p lt p lle O pois p lt O p lle O pois p lt o Op lle O pois p lt TURKU UNIVERSITY OF APPLIED SCIENCES BACHELOR S THESIS Donald Egbenyon 46 Logout Toiminta Koti Osoitekirja Asetukset Ominaisuudet Pikavalinta Toimintan pp imet Identiteetti 1 Identiteetti 2 Identiteetti 3 Identiteetti 4 Identiteetti 5 Identiteetti 6 Identiteetti 7 Identiteetti 8 Identiteetti 9 Identiteetti 10 Identiteetti 11 Identiteetti 12 Action URL Lis asetukset Certificates Ohjelmiston p ivitys Tilatiedot Tietoa j rjestelm st Loki SIP Trace DNS Cache Subscriptions PCAP Trace Lis asetukset Network Behavior Audio Quality of Service Type of Service TOS SIP Type of Service TOS Diffserv VLAN VLAN Id 0 4095 VLAN Priority 0 7 Un Tag VLAN traffic to from specific switch ports Verkon Portti VLAN Id 0 4095 VLAN Priority 0 7 PC Portti VLAN Id 0 4095 VLAN Priority 0 7 IEEE 8021x Authentication K ytt j Salasana Ignore security advices Use hidden tags Allow CSTA control Empty client cert Filter Packets from Registrar Authentication for SIP Reboot Authentication for SIP Check Sync Administraattorimoodi Administraattorin salasana Administraa
16. Bachelor s Thesis UAS Degree Program Information Technology Specialization Data Communication amp Networking 2011 Donald Egbenyon Implementing QoS for VoIP ina Local Area Network LAN N TURUN AMMATTIKORKEAKOULU TURKU UNIVERSITY OF APPLIED SCIENCES BACHELOR S THESIS ABSTRACT TURKU UNIVERSITY OF APPLIED SCIENCES Degree Programme Information Technology November 2011 54 Instructors Harri Waltari Ossi Vaananen Quality of Service QoS is being deployed by most VoIP service providers today in order for such enterprises to efficiently use its bandwidth in an integrated network These service providers expect their respective clients to implement QoS on their LAN in order to maximize the bandwidth resources at their disposal and ensure a flawless and high quality VoIP for its use With the necessary equipment in the LAN the network administrator with the required knowledge in QoS design and implementation can configure the various routers and switches to give priority to voice in the LAN This thesis focuses on the implementation of Quality of Service QoS in a LAN that extends over an IP MPLS network QoS was configured on all the switches in a particular network in order for priority to be given to the Voice packets going through this network The aim is to show that although QoS is not so important in a LAN within a Region it becomes important when that LAN extends to other Regions over an IP MPLS network
17. D SCIENCES BACHELOR S THESIS Donald Egbenyon 36 Table 6 2 192 168 248 51 to 192 168 248 44 QoS Baseline Ping results in Region 1 Packet Size WithOut QoS Policy With QoS Policy 160 bytes min avg max ms min avg max ms Extended ping 1 1 13 1007 1 4 34 Extended ping 2 1 3 25 1 3 17 Extended ping 3 1 3 16 1 3 16 Success rate Success rate Extended ping 1 100 100 Extended ping 2 100 100 Extended ping 3 100 100 The results in Table 6 1 and 6 2 do not clearly show any proof that QoS will be beneficial when implemented in a LAN since there is little difference between the min avg max time as this LAN does not extend over an MPLS Region 1 and Region 2 are on the same LAN but this LAN extends over an IP MPLS as can be seen in the network topology in Figure 4 2 Since the LAN extends over an IP MPLS the extended ping before QoS is implemented and after the implementation the effect of QoS implementation is clearly shown TURKU UNIVERSITY OF APPLIED SCIENCES BACHELOR S THESIS Donald Egbenyon 37 Table 6 3 192 168 248 51 to 193 142 250 3 QoS Baseline Ping results from Region 1 to Region 2 Packet Size WithOut QoS Policy With QoS Policy 160 bytes min avg max ms min avg max ms Extended ping 1 8 16 26 1 8 17 Extended ping 2 8 16 26 1 8 17 Extended ping 3 8 16 33 1 8 25 Success rate Success rate
18. Layer Application Data Protocol sip tcp TLSv1 Record Layer Application Data Protocol sip tcp 0000 0014 a9 c3 55 ff fO de f1 Of c7 6d 08 00 45 00 0010 04 22 Oe 40 40 00 80 06 00 00 c0 a8 f8 79 3e 47 0020 6a fb do 85 13 c5 bl 64 25 10 1c 52 fb 27 50 18 0030 01 00 66 79 00 00 17 03 01 00 20 9b 86 f3 87 b3 0040 3e f6 ae Oe b3 11 df 9c 26 c 11 11 d9 f2 01 5f ifanen Ad d af he a IN d 30 E hO dh 17 A MI A AA File C Documents and Settings Ehis My Docum Packets 1249 Displayed 1249 Marked 0 Load time 0 00 500 Profile Default Figure 4 3 Analysing TLSv1 protocol From Figure 4 3 when frame number 40 is also analysed it was noticed that the protocol used by the client to send an acknowledgment is the TLSv1 protocol When there is communication between a TLS client and a TLS server the hello reguest message is optional but in this case it was sent by the client to the server and the session had already been established It can also be seen that the communication between the client and the server is a secured communication because Secure Socket Layer is being used for this communication since it is a TLSv1 protocol When the session between the client VIP soft phone in the computer has been established to the other client VIP soft phone in the handset the protocol was changed to a UDP This is because VolP communication cannot use a TCP protocol for communication This can be seen in F
19. NIVERSITY OF APPLIED SCIENCES BACHELOR S THESIS Donald Egbenyon Table 4 1 Call Picked by VIP softphone on handset 21 Source IP Source Dest Dest IP address Port Port address 192 168 248 121 53381 11 gt 5061 67 71 106 251 53381 n13 5061 53397 gt 443 67 71 106 250 53397 4 443 49154 1 2 62306 67 71 106 244 53381 nn1 1 5061 62 71 106 251 49154 17 62306 62 71 106 244 Ports Used Ports Used 53381 5061 53397 443 49154 62306 Step 2 With the results gathered from the first call made from the laptop it cannot be ascertained if these ports are static ports or if the VIP soft phone uses a range of dynamic ports Another call was made but this time it was picked by the soft phone on a laptop After analysing the data captured with Wireshark it was noticed that the ports changed slightly from the ports gathered from the previous result The table below shows that fewer ports were used in this communication compared with the previous communication Table 4 2 Call picked by VIP softphone on Laptop Source IP Source Dest Dest IP address Port Port address 192 168 248 121 49269 5061 67 71 106 250 49269 11 5061 49154 _ 56674 67 71 106 242 49154 a 56674 Ports Used Ports Used 49269 5061 49154 56674 TURK
20. S cvoice foundation learning guide 4th ed Indianapolis Ind London Cisco Pearson Education distributor 25 A sample configuration of AutoQoS for VoIP 2003 Retrieved July 2011 from http www avaya com uk emea en us resource assets applicationnotes autogos pdf 26 Cisco Systems Inc 2003 Cisco AutoQoS whitepaper Retrieved July 2011 from http www cisco com en US tech tk543 tk759 technologies white paper09186a00801 348bc shtml TURKU UNIVERSITY OF APPLIED SCIENCES BACHELOR S THESIS Donald Egbenyon Appendix SIP RTP TURKU UNIVERSITY OF APPLIED SCIENCES BACHELOR S THESIS Donald Egbenyon 44 w 45 LSS 5 ja aN EN A MAE Rd Show Diversion O p lle pois p lt Tietoa j rjestelm st A a wi Use NAPTR on SIP URIs Op lle pois p lt O oki A x Si Encode display name p lle pois p lt race de Update Called Party ID Op lle Opois p lt O eee Voice Quality Report Collector PCAP Trace 7 a i Check SDP Version p lle pois p lt Muisti Asetukset Minibrowser K ytt ohje XML NOTIFY Support Op lle O pois p lt RTP RTCP snom 2000 2010 snom AG Dynaaminen RTP portti alku Dynaaminen RTP portti loppu DTMF payload tyyppi RTCP Tuki RTP Keepalive Enhanced JB JB Min JB Max JB Shrink Multicast Multicast Support 1 IP Address 2 IP Address 3 IP Address 4 IP Address 5 IP
21. U UNIVERSITY OF APPLIED SCIENCES BACHELOR S THESIS Donald Egbenyon 22 From the table above it was established that fewer ports are used for a SIP communication to a soft phone in a laptop compared with more ports that are used when the communication ends on a soft phone installed on a handset as can be seen in Table 4 1 This process of making calls to the handset or laptop where the VIP soft phone is installed was repeated several times to know if the same ports used will be the same as the one earlier gathered or not After studying different data captured by Wireshark it became obvious that different ports are used whenever a call is made between users 4 1 Deductions from analysed data So from the data gathered using Wireshark it became obvious that the range of RTP UDP dynamic ports from 49152 65535 is being used by VIP soft phone to communicate with the server These range ports are used for media reguests sent from the client to the server They are also used for inbound and outbound media transfer through the firewall From the data analysed SIP uses signal port 5061 for communication and authentication between the client and the server It is also used for outbound communication from the client to the server through the firewall In addition it was observed that HTTPS TCP port 443 is used by the client to send SIP traffic to the server Most times this port is also used by client connecting to the server fo
22. VIP soft phone or SNOM hard phone in the network since it recognises only Cisco devices These are the steps used when using MQC to configure a network 1 Class Map Defining the class of traffic needed Each class is defined using a class map command 2 Policy Map The QoS policies for the classes are defined using policy map command This states what will be done to the traffic defined in the class map 3 Service Policy This attaches the configured policy to an interface using the service policy command Without attaching the policy the class map and policy map configured will not be used by the device 5 2 Lab test Usually in an enterprise network it is wrong to test configurations in a real time environment It is important for the configurations to be fine tuned in a lab environment before it is implemented in the network in order not to disrupt the network traffic The MQC was tried initially in a lab environment before it was implemented gradually in the network The lab topology can be seen below TURKU UNIVERSITY OF APPLIED SCIENCES BACHELOR S THESIS Donald Egbenyon 27 Lab Testing Topology m X Ol me gt gt i SEI Mutlayer Sten Figure 5 1 Lab Logical Topology The lab network is made up of two switches Cisco C2960 and a C4503 switch connected with a crossover cable The configuration used was designed with an access list that has a ping command attached Part of the
23. action policed dscp transmit class DOM VOIP CONTROL set ip dscp af31 police 1000000 16000 exceed action policed dscp transmit TURKU UNIVERSITY OF APPLIED SCIENCES BACHELOR S THESIS Donald Egbenyon 48 Turning QoS on mls qos COS setting mls qos map cos dscp 0 8 16 26 34 46 48 56 Attaching to an interface int gi0 48 service policy input DOM QOS TRAFFIC Voice RTP and Voice Control access list ip access list extended VOIP RTP permit udp any any range 49152 65535 permit ip any any precedence 5 permit ip any any dscp ef permit ip any any dscp cs5 ip access list extended VOIP CONTROL permit tcp any any eq 5061 permit tcp any any eq 443 TURKU UNIVERSITY OF APPLIED SCIENCES BACHELOR S THESIS Donald Egbenyon
24. cations at the same time without so many difficulties experienced in circuit switch technology In the case of VoIP the major factors that will affect its quality are packet loss and packet delay Lost packets Usually when data is being sent over an IP network some packets can be lost Using TCP protocol the lost packet can be resent but in the case of VoIP it does use UDP So any packet lost cannot be retransmitted When a packet is lost it brings about voice clipping and skips There is a standard industry codec used to correct up to 30ms of lost voice in most Digital Signal Processor found on routers and switches Delay Packet Latency Packet delay is the time it takes a packet to reach the receiving end of an endpoint after it has been transmitted from the source This is called end to end delay It consists of two components fixed network delay and variable delay Packet delay can cause degradation of voice quality due to the end to end voice latency or packet loss if the delay is variable 6 The end to end voice latency must not be longer than 250ms because it will make the conversation sound TURKU UNIVERSITY OF APPLIED SCIENCES BACHELOR S THESIS Donald Egbenyon like two parties talking on a CB radio If latency must be taken care of then we have to know the main causes of latency They are codecs queuing waiting for packets being transmitted serialization jitter buffer and others Queuing waiting for packets being tran
25. configuration can be seen below This configuration shows just the access list Voice RTP and Voice Control access list ip access list extended VOIP RTP permit udp any any range 49152 65535 permit ip any any precedence 5 permit ip any any dscp ef permit ip any any dscp cs5 permit ip any any echo permit ip any any echo reply ip access list extended VOIP CONTROL TURKU UNIVERSITY OF APPLIED SCIENCES BACHELOR S THESIS Donald Egbenyon 28 permit tcp any any eq 5061 permit tcp any any eq 443 Echo and echo reply was added to the access list in order to determine if ping is given expedited forwarding or not It was noticed from the results gathered from the lab work that there was a little difference in the ping result when QoS was implemented on one of the switches This will be explained more clearly in the next chapter when QoS has been implemented in the network 5 3 Configuration used It will also be good to have an overview of what the logical network looks like in order to have a grasp of the network and the way packets flow from the different offices through the LAN This is also necessary since QoS implementation will be complex or simple depending on the traffic patterns and types in the network These switches and router make up the network topology This can be seen in Figure 5 2 below TURKU UNIVERSITY OF APPLIED SCIENCES BACHELOR S THESIS Donald Egbenyon 29 Switches n
26. e is congestion in the network based on the traffic significance QoS is not something that will be configured on a router or switch rather it is a term that refers to a wide variety of mechanism used to influence traffic patterns on a network 21 It gives network administrators the ability to give some traffic more priority over others 2 1 Models of QoS The purpose of QoS usage is to make sure that minimum bandwidth is guaranteed for identified traffic jitter and latency is being controlled and packet loss is improved This can be carried out using several means either QoS congestion management congestion avoidance or policing and traffic shaping The means chosen depends largely on the goal of the network administrator QoS can be divided into three different models These models describes a set of end to end QoS capabilities 14 In order to facilitate end to end QoS on an IP network the Internet Engineering Task Force IETF defined two models Integrated Services IntServ and Differentiated Services DiffServ 5 A default model that comes with all networking devices is the Best Effort model It does not require any QoS configuration The IntServ model follows the end to end signalling process whereby the end hosts tells the network their QoS needs in advance while DiffServ follows the provisioned QoS model whereby the network elements set up multiple classes of traffic with different degrees of QoS needs IntServ provides high QoS
27. e or IP precedence value This is possible because DSCP is backward compatible with IP precedence value IP precedence values range from 0 to 7 while DSCP values range from 0 to 63 This can be seen in the table below TURKU UNIVERSITY OF APPLIED SCIENCES BACHELOR S THESIS Donald Egbenyon Table 2 1 Comparing IP Precedence and DSCP values IP Precedence and DiffServ Precedence DiffServ DSCP 7 111Xxxxx Class selector CS7 111000xx 56 6 110xxxxx Class selector 6 CS6 10000xx 48 Expedited Forwarding EF 101110xx 46 5 101xxxxx Class selector 5 CS5 101000xx 40 Assured Forwarding 4 AF4 100dd0xx 34 36 38 4 100xxxxx Class selector 4 CS4 100000xx 32 Assured Forwarding 3 AF3 011dd0xx 26 28 30 3 011Xxxxx Class selector 3 CS3 011000xx 24 Assured Forwarding 2 AF2 010dd0xx 18 20 22 2 010XxXxxxx Class selector 2 CS2 010000x 16 Assured Forwarding 1 AF1 001dd0xx 10 12 14 1 001xxxxx Class selector 1 CS1 001000xx 8 O O00xxxxx Best Effort BE 000000xx 0 TURKU UNIVERSITY OF APPLIED SCIENCES BACHELOR S THESIS Donald Egbenyon IP precedence 7 or DSCP 56 is reserved for network use The DSCP values used in network configuration range from 0 to 48 while the IP precedence values used range from 0 to 7 Any of the DSCP values can be used to assign priority to a traffic flow TURKU UNIVERSITY OF APPLIED SCIENCES BACHELOR S THESIS Donald Egbenyon 3 QoS for Voice Voice packets have minimal ne
28. ecedence DSCP and SNOM ToS Value 18 Table 4 1 Call Picked by VIP softphone on handset 41 44 10 10 11 11 12 12 17 18 19 20 27 29 34 39 13 21 Table 4 2 Call picked by VIP softphone on Laptop Table 5 1 Devices in the network configured Table 5 2 Summary of ACL Restrictions Table 5 3 Syntax Description Table 6 1 192 168 248 51 to 192 168 248 235 QoS Baseline Ping results in Region 1 Table 6 2 192 168 248 51 to 192 168 248 44 QoS Baseline Ping results in Region 1 Table 6 3 192 168 248 51 to 193 142 250 3 QoS Baseline Ping results from Region 1 to Region 2 Table 6 4 192 168 248 51 to 193 142 250 8 QoS Baseline Ping results for Region 1 to Region 2 21 30 32 32 35 36 37 37 ACRONYMS ABBREVIATIONS AND SYMBOLS SNOM SIP QoS LAN VolP IP IntServ DiffServ IETF CoS ToS DSCP PSTN VolP TCP UDP RTP VIP HTTPS STCP PHB ACE SNOM Technology AG Session Initiation Protocol Quality Of Service Local Area Network Voice Over internet protocol Internet Protocol Integrated Services Differentiated Services Internet Engineering Task Force Class of Service Type of Service Differentiated Service Code Point Public Switch Telephony Network Voice Over Internet Protocol Transmission Control Protocol User Datagram Protocol Real time Transport protocol Sonera VIP softphone Secured Hypertext Transport Protocol Secured Transmission Control Protocol
29. eds for delay Voice packets are delay sensitive and a little delay in the transmitting of voice packets can cause so much discomfort to a user For a long time the Public Switched Telephone Network PSTN used to carry voice traffic from one user caller to another user receiver The PSTN uses a circuit switched network This means that when the caller decides to call the receiver a circuit is open in the network between the caller and the receiver until the call is aborted It guarantees that the caller can have a high quality call for as long as the call last This circuit switched network brings some difficulties because no other call can go through the already open circuit until it is closed Also it is expensive for the user since the user pays for the open circuit In order to use a less expensive means of carrying voices over a network the voice packet was designed to be carried as a packet over an Internet Protocol IP This gave birth to Voice over IP VoIP The major question in the minds of sceptics was if this new technology will be able to guarantee the same Quality of Service QoS that is required in voice communication This is because the voice packet is not circuit switched rather it is packet switched over the network Packet switching entails the packet to be broken into fragments from the sender device and then reassembles before it gets to the receiver Packet switching technology is the ability to carry several voice communi
30. esults gathered before implementing QoS and after the implementation 6 1 1 QoS Baseline Before applying any QoS configurations an extended ping from one of the switches 192 168 248 51 is sent to all the switches in the network in the three regions that make up the LAN 3 times for each switch Below is a sample picture of the ping to one switch in region 2 office Figure 6 1 Region 2 switch TURKU UNIVERSITY OF APPLIED SCIENCES BACHELOR S THESIS Donald Egbenyon 6 1 2 After configuring QoS the extended ping test was repeated on all the switches again by adding echo and echo reply to the access list that permits Expedited Forwarding Then a comparison was made between the ping before the QoS was configured on the network and after it was configured The table below sums up the results The results from four switches two from the switches in Region 1 and two from those in Region 2 QoS Baseline Results are used in this project only to explain the effects of QoS configuration Table 6 1 192 168 248 51 to 192 168 248 235 QoS Baseline Ping results in Region 1 Packet Size WithOut QoS Policy With QoS Policy 160 bytes min avg max ms min avg max ms Extended ping 1 1 3 9 1 2 9 Extended ping 2 1 3 25 1 3 17 Extended ping 3 1 2 9 1 3 17 Success rate Success rate Extended ping 1 100 100 Extended ping 2 100 100 Extended ping 3 100 100 TURKU UNIVERSITY OF APPLIE
31. ev 38931 from trunk 1 6 DER File Edt View Go Capture Analyze Statistics Telephony Tools Internals Help EELE SBAXSS Ve odFL FSi QAQAQH E Filter Expression No Time Source Destination Protocol Length Info n 32 3 452500 192 108 248 95 192 108 248 121 Lome 4ECNO PIN reply 10 UXxUVUL Seq 33 5 452072 Ibm 9e 78 ee Broadcast ARP 60 who has 192 168 249 137 Tell 192 34 5 544454 62 71 106 251 192 168 248 121 TLSV1 1434 Application Data 35 5 544548 62 71 106 251 192 168 248 121 TCP 1434 TCP segment of a reassembled PDU 544593 a 559837 192 168 248 121 TCP 60 sip tls gt 53381 ACK Seq 3435 Ack r Spanning tree for br S1TP 60 RST Root 8192 1 00 14 39 c3155 c 13 23 j 1 CH 45 6 413645 62 71 106 251 192 168 248 121 TCP 60 sip tls gt 53381 ACK Seq 3435 Ack 46 6 413675 62 71 106 251 192 168 248 121 TLSVL 128 Application Data Application Data i 47 6 433201 192 168 248 121 192 168 248 55 ICMP 74 Echo ping request id 0x0001 seq 48 6 439048 192 168 248 55 192 168 248 121 ICMP 74 Echo ping reply id 0x0001 seq 49 6 501411 Vmware ac 00 0d Broadcast ARP 60 who has 192 168 248 397 Tell 192 16 50 6 505384 192 168 248 12 192 168 249 255 gt amp Frame 1 60 bytes on wire 480 bits 60 bytes captured 480 bits amp Ethernet II Src wistron 1e f2 68 00 1f 16 1e f2 68 Dst Broadcast ff ff ff ff ff ff amp Address Re
32. g voice traffic or not All traffic is given the required attention by the networking devices and delivered to the required destination 6 2 Conclusion Since implementing QoS on a LAN might have minimal effects on the network when such a LAN is within the same region it is obvious that if that LAN extends over a WAN then implementing QoS in a network will be of paramount importance since the network can give preferential treatment to set traffic like VoIP The QoS configuration on such a converged network brings low latency low jitter and high availability to VoIP traffic on such a network Moreover it can be seen clearly that in a network with diverse switches or routers different QoS configurations have to be designed and implemented on such a network Furthermore DiffServ model of QoS is the most scalable and better model for implementing QoS on a network since it allows the use of Modular QoS CLI to be used on different networking devices TURKU UNIVERSITY OF APPLIED SCIENCES BACHELOR S THESIS Donald Egbenyon 41 REFERENCES 1 Alvarez S amp Cisco Systems 2006 QoS for IP Indianapolis Cisco Press 2 Cisco Systems Inc 2001 Cisco IP telephony QoS design guide pp 92 106 San Jose CA Cisco Systems Inc 3 Cisco Systems Inc 2005 Configuring QoS Catalyst 2960 series switch cisco IOS software configuration guide 12 2 25 FX ed pp 487 573 San Jose CA Cisco Systems Inc 4 Cisco Systems
33. he network it is paramount that different configurations need to be designed and loaded on these different switches There are some cases where a particular configuration will fit into different switches The Cisco IOS has different limitations on different Cisco platforms Each configuration starts with a class map a policy map access list and then attaching the policy to an interface either a physical interface or a port channel Some sample configurations that were suitable for C2950 and C2960 can be seen in the Appendix The main reason different configuration will be needed is because every switch has different IOS models and these models are optimised for some specific purposes Moreover since technology is evolving the manufacturer of such device cannot integrate all features into one switch model since that will be difficult to do TURKU UNIVERSITY OF APPLIED SCIENCES BACHELOR S THESIS Donald Egbenyon 31 The switches were configured in this network but the routers were not configured since they are the property of the service provider installed in the company network The routers have already been configured by the service provider since they are the provider s edge PE device It is worthy of note that this topology shows an MPLS Layer 2 VPNs which is a Layer 2 switched solution This kind of approach allows for a separation of the customer network from the provider s network thus there is no route exchange between the customer
34. igned and implemented it has the ability to save bandwidth and the equipment used on such a network The heart of the converged network is VoIP VoIP is the process of converting analogue audio signals to digital signals that can be transmitted over the Internet It provides voice and telephony services over an IP network The discovery of VoIP changed the paradigm of IP over everything to everything over IP 5 This arrival of VoIP showed a flaw in the way IP carries packets Voice packets will not wait to be delivered at the time the network chooses but it specifies a particular time of delivery which the network must abide to This means that the best effort delivery model of the IP network cannot be tolerated by voice packets The network has to be configured to give more priority to some packets while giving less priority to other packets This means that Quality of Service QoS has to be configured on a network in order for different priority to be given to traffic TURKU UNIVERSITY OF APPLIED SCIENCES BACHELOR S THESIS Donald Egbenyon So because converged networks have taken a firm hold on the communication industry most service providers of VoIP are encouraging the use of a converged network instead of separate VLANs for the voice data and video packets In order for these voice and data packets to be carried in the same VLAN the network must be properly configured with the right priorities given to voice data and video
35. igure 4 4 below TURKU UNIVERSITY OF APPLIED SCIENCES BACHELOR S THESIS Donald Egbenyon 20 7 capturet pcap Wireshark 1 6 2 SVN Rev 38931 from trunk 1 6 File Edt View Go Capture Analyze Statistics Telephony Tools Internals Help VUSON EHKXZSS 46 99T2EBIAAGAROIWVEMAIN Expression Source Destination Protocol Length Info 1148 36 945271 192 168 248 134 192 168 249 255 NBNS 92 Name query NB WPAD lt 00 gt i 773 192 168 248 121 62 71 244 74 Source port 49155 Destination port 62307 1153 37 513897 192 168 248 121 192 168 248 55 74 Echo ping request id 0x0001 seq 15042 49722 TA 1154 37 514815 192 168 248 55 192 168 248 121 74 Echo ping reply id 0x0001 seq 15042 49722 1 4 1 1159 37 738281 192 168 248 134 192 168 249 255 NBNS 92 Name query NB WPAD lt 00 gt 1160 37 797264 Canon 0b b0 c9 Broadcast ARP 60 Gratuitous ARP for 169 254 94 59 Reguest 1161 37 988240 Ibm 9e 78 ee Broadcast ARP 60 who has 192 168 249 937 Tell 192 168 248 19 1162 38 314145 i 96 07 1 Spanning tree for hr s f 60 RST Root 8192 1 00 14 39 c3155 c0 Cost 1163 3R 4RR339 197 168 74R 134 197 168 249 755 NANS 97 Name query NR WPAN lt 00 gt Frame 1152 74 bytes on wire 592 bits 74 bytes captured 592 bits 8 Ethernet II Src wistront_of c7 6d fO de fl 0f c7 6d Dst Cisco c3 55 ff 00 14 a9 c3 55 FF amp 1 User Datagram Protocol Src Port 49155 49155 Dst Port 62307 62307 F Source
36. nation address and application Classification is the most fundamental QoS building block Without it all packets are treated the same Usually it should take place at the network edge maybe in the wiring closet or in the IP phones or in voice endpoints The packets can be marked as important using Layer 2 Class of Service CoS setting on the User Priority bits of the 802 1p portion of the 802 1Q header or on the Differentiated Services Code TURKU UNIVERSITY OF APPLIED SCIENCES BACHELOR S THESIS Donald Egbenyon 10 Point DSCP IP precedence bits in the Type of Service ToS byte of the IPv4 or IPv6 header as shown in Figure 3 1 3 2 3 3 and 3 4 The diagram in Figure 3 1 shows the ToS field in IPv4 and IPv6 headers while the diagram in Figure 3 2 and Figure 3 3 shows the original IPv4 ToS field T f Traff Ver4 IHL Peo Total Length vere ae Flow Label Service Class Identification Flags Frag Offset MW Payload Length Next Hdr Mist Protocol Header Checksum Live Source Address Source Address Destination Address Destination Address Figure 3 1 IPv4 and IPv6 headers 5 This figure shows the ToS field before it was renamed to be the DS field oO a Bits IPv4 TOS Byte Precedence DTR Bits RFC 1122 RFC 1349 Zero Bits 0 2 IP Precedence Defined Bits 3 6 The Type of Service Defined 111 Network Control 0000 all normal 110 Internetwork Control 1000 minimize delay 101
37. oIP requires two types of protocols a media protocol and a signalling protocol The media protocol manages the transmission of voice packets over an IP networks Examples are Real time Transport Protocol RTP Real time Transport Control Protocol RTCP and Secure Real time Transport Protocol SRTP The signalling protocol manages call setup and call tear down Examples of signalling protocols are H 323 SIP MGCP Skype etc The signalling protocol uses Transmission Control Protocol TCP for its transmission while the Media protocol uses User Datagram Protocol UDP as its transmission b Signal Protocol A signal protocol is a type of protocol used to identify signalling encapsulation It is used to identify the state of connections between telephones or VoIP 19 VoIP has various standards that are being used for signalling In the case of Sonera they decided to use the SIP standard In order for SIP to be appreciated it will be better to have a good understanding of some other standard used as signal protocol such as H 323 c H 323 H 323 is an ITU recommended packet based multimedia communication system that defines a distributed architecture for creating multimedia applications like VoIP It is seen as an umbrella standard that covers other sub protocols like H 225 and H 245 TURKU UNIVERSITY OF APPLIED SCIENCES BACHELOR S THESIS Donald Egbenyon 16 connected to signalling and call setup Its strength lies in its ability to
38. oning means making the network available for all the traffic that uses the network 15 The network provisioning tools ensure that the needed bandwidth is accurately calculated for voice traffic data traffic video applications and link management like routing protocols It is worthy of note that the bandwidth used by the voice data and other applications should not exceed 75 of the provisioned bandwidth This will ensure that the remaining bandwidth of 25 is used for management purposes like routing protocols VOIP bandwidth calculations and others TURKU UNIVERSITY OF APPLIED SCIENCES BACHELOR S THESIS Donald Egbenyon 15 4 Gathering Information The right information is vital when a network needs to be configured for efficiency A network engineer ought to know the required switches or routers that need configuration and which of these do not require configurations Also the engineer ought to know the ports to be configured and the interfaces to apply such configurations In order for this to be done network traffic analysis tools like Wireshark will be needed to gather the right information from the network In order for this work to proceed there are some terms that need to be explained since it was noticed that different protocols are being used from the Wireshark data What is a signal protocol This will be answered by discussing Signal Protocol and some standard Signal protocols used today They are a VolP Protocols V
39. packet Type of Service TOS byte in the IP header is marked in order for the packets to be divided into different classes which forms the aggregate behaviours 5 Differentiated Services Code Point DSCP is a 6 bit bit pattern in the IPv4 TOS Octet or the IPv6 traffic class Octet 5 DSCP supports up to 64 aggregates or classes and all classification and QoS in the DiffServ model revolves around the DSCP Though classification is carried in the IP header it can also be carried in the Layer 2 frame These special bits in the Layer 2 frame or Layer 3 packet are described below A Prioritization bits in Layer 2 frames Layer 2 Inter Switch Link ISL frame headers carries the IEEE 802 1p class of service COS in the 1 byte User field while the Layer 2 802 1Q frame headers carries the COS value in a 2 byte Tag Control Information field The Tag Control Information TAG field carries the CoS value in the 3 most important bits which is normally called User Priority bits in a Layer 2 802 1Q header frame Layer 2 CoS values range from O for low priority to 7 for high priority The figure below shows the 2 byte TAG TURKU UNIVERSITY OF APPLIED SCIENCES BACHELOR S THESIS Donald Egbenyon Three bits used for CoS User Priority Figure 2 1 Overview of the TAG and User Priority in a Layer 2 802 10 frame 26 B Prioritization bits in Layer 3 packets Layer 3 IP packets can carry either Differentiated Services Code Point DSCP valu
40. port 49155 49155 Destination port 62307 62307 Length 40 TH Checksum 0x6297 validation disabled g Data 32 bytes Data 81 9000721c90631e290e6020000000000006ac800000003 Length 32 Tijoooo 00 14 a9 c3 55 fF fo de f1 OF c7 6d 08 00 45 00 FI 00 3c 11 e9 40 00 80 11 00 00 c0 a8 f8 79 3e 47 6a f4 cO 03 f3 63 00 28 62 97 81 c9 00 07 21 c9 Ty 06 31 e2 90 e6 02 00 00 00 00 00 00 a c8 00 00 00 03 00 04 80 c4 0002 6F 98 nae ee J O File C Documents and Settings Ehis My Docum Packets 1249 Displayed 1249 Marked 0 Load time 0 00 500 Profile Default L Figure 4 4 Using UDP for VolP communication Usually SIP communication can both use UDP or TCP as the transport protocol for transfering data Initially UDP was the only protocol allowed in RFC2543 but in RFC3261 both TCP and UDP were given the permission to be used as the data transfer transport protocol Today UDP is still the most used in most SIP communications In this communication as can be seen in the figure above the UDP protocol was used until the call was ended by the users The last UDP protocol used can be seen in frame 1110 and after the call was ended a TCP protocol was used to eventually bring the session to an end After analysing the Wireshark captured results it became obvious that a range of protocols and ports were used in the communication The table below shows the ports used in the communication from the Wireshark data TURKU U
41. r SIP communication outside an intranet Lastly all communication for transferring important data uses TLSv1 This makes it difficult for hackers to hack the information in the server or intercept the communication between several users TURKU UNIVERSITY OF APPLIED SCIENCES BACHELOR S THESIS Donald Egbenyon 23 5 QoS Implementation QoS implementation as was earlier mentioned can be a simple or complex task depending on several factors like the QoS features offered by the networking devices the traffic types and pattern in the network and level of control that need to be exercised over incoming and outgoing traffic The network engineer cannot even be able to exercise so much control over the network traffic beyond the QoS features in the networking devices 5 1 QoS Model used There are diverse models for implementing QoS on a network As stated earlier though IntServ Best effort and DiffServ can be used to implement QoS the best and most scalable model for implementing QoS in a network is the DiffServ model The DiffServ model allows network traffic to be broken down into small flows for appropriate marking This small flow is called a class Thus the network recognises traffic as a class instead of the network receiving specific QoS request from an application The devices along the path of the flow are able to recognise the flow because these flows are marked So the marked flow is given appropriate treatment by the va
42. rement of time Since the traffic in this particular network was divided into two classes the voice and others by the network administrator it is important that LLQ and traffic policing is used to implement the various configurations in the different switches How then is QoS implemented in a network It starts with preparation TURKU UNIVERSITY OF APPLIED SCIENCES BACHELOR S THESIS Donald Egbenyon 25 5 1 1 Preparing to Implement the QoS model a Identifying types of traffic and their requirement Using Wireshark the traffic required can be identified In this scenario Wireshark was used to confirm the range of ports that are to be configured In addition it is necessary to know the business importance of all traffic in a network b Dividing traffics into classes The identified traffic is divided into two classes Voice and Others This is specified by the network administrator for the network In this scenario the traffic was divided into two classes Voice and Others The Voice class is given low latency while Others will be configured with guaranteed delivery c Defining QoS policies for each class Defining the QoS policy involves one or more of the following activities setting a minimum bandwidth guarantee setting a maximum bandwidth limit assigning a priority to each class and using QoS technology to manage congestion 5 1 2 Implementing the chosen QoS model Since it has been established that the QoS model
43. rious devices on the network In the previous chapter the DiffServ field was shown with some diagrams to clearly explain what it really looks like in IPv4 or IPv6 When the packet has been properly marked and identified by the router or switch it is given special treatment called Per Hop Behaviour PHB by these devices PHB identifies how the packets are treated at each hop There are three standardized PHB currently in use a Default PHB This uses best effort forwarding to forward packets b Expedited Forwarding EF It guarantees that each DiffServ node gives low delay low jitter and low loss to any packet marked with EF It is used majorly for Real time Transport Protocol RTP applications like voice and video c Assured Forwarding AF It gives lower level functions compared with EF It is used mostly for mission critical applications or any application that are not too sensitive to delay but want assurance that the packets will be delivered by the network TURKU UNIVERSITY OF APPLIED SCIENCES BACHELOR S THESIS Donald Egbenyon 24 There are several tools used in implementing QoS These are some tools used in QoS implementation 1 Congestion Management Queuing is meant to accommodate temporary congestion on an interface of a network device by storing the excess packets until there is enough bandwidth to forward the packets Sometimes some packets are dropped when the queue depth is full Congestion management allow
44. s field in the IP header 4 4 8 16 1 3 13 6 8 16 32 32 hits VERS HLEN TOS Total Flags Frag TTL Protocol Header SA DA IP Length Offset Checksum Options Figure 3 5 TOS 8 bits field 18 The ToS field was renamed to the Differentiated Services field 6 bits of the DS field is used as the code points for selecting the per hop behaviour PHB while the last 2 bits are unused just like the original ToS field Figure 3 6 gives a clear idea of what it looks like Bits 3 A 1 1 1 1 TOS with Reliabilit BZ IP Precedence Precedence Delay Threughput eliabili Cost MA Bits 6 2 SS DifServ Code Point Unused DSCP Figure 3 6 Comparing ToS to IP Precedence and DSCP 18 Initially it was established that the ToS field was not used by most manufacturers of networking devices and this is the reason why it was renamed to the DS field Today SNOM decides to use the ToS field in making the SNOM hard phone According to SNOM hard phones manufacturers the value 160 means IP precedence 5 or DSCP Class Selector 5 CS5 16 It is well explained in the table below TURKU UNIVERSITY OF APPLIED SCIENCES BACHELOR S THESIS Donald Egbenyon 13 Table 3 1 Comparing IP Precedence DSCP and SNOM ToS Value 18 IP Prec IP DSCP Class DSCP DSCP ToS AA Hex Dec value SNO M 0 000 Best effort 000000 0x00 0 0 1 001 CS 1 001000 0x08 8 32 AF11 Low 001010 Ox0A 1
45. s the administrator to control congestion by determining how and when the queue depth is full There are several ways this can be done These are some ways congestion management can be implemented a Priority Queuing PQ This allows the administrator to give priority to certain traffic while allowing other to be dropped when the queue depths are full b Custom Queuing CQ This allows the administrator to reserve queue space in the router or switch buffer for all traffic types c Weighted Fair Queuing WFQ This allows the sharing of bandwidth with prioritization given to some traffic d Class based Weighted Fair Queuing CBWFQ This extends the functionality of WFQ to provide support for user defined classed e Low Latency Queuing LLQ This is a combination of CBWFQ and PQ It is able to give traffic that requires low delay the required bandwidth it needs while also giving data the needed bandwidth It solves the starvation problem associated with PQ 2 Traffic shaping and traffic Policing Traffic shaping and policing are mechanisms used to control the rate of traffic The main difference between them depends on the terms of implementation While traffic policing drops excess traffic or remarks the traffic in order to control traffic flow within a specific rate limit without introducing any delay to traffic traffic shaping retains excess traffic in a queue and then schedules such traffic for later transmission over an inc
46. se it as its signalling protocol instead of H 323 SIP has two kinds of ports 5060 and 5061 Port 5060 is used for un encrypted communication while port 5061 is used primarily for an encrypted communication over the Internet The encrypted communication is carried over Transport Layer Security TLS Transport Layer Security TLS is a cryptographic protocol that provides communication security over the Internet 20 Its predecessor is Secure Socket Layer SSL It encrypts the segments of network connections above the Transport Layer using asymmetric cryptography thus providing channel oriented security Therefore HTTPS is a combination of HTTP with SSL TLS and this is what SIP uses to provide a secure connection TURKU UNIVERSITY OF APPLIED SCIENCES BACHELOR S THESIS Donald Egbenyon 17 Wireshark was used in this research to monitor the flow of traffic for some days in order to ascertain the ports used for communicating between the client computer or softphone and the server In order for this to be done a phone call was made from the VIP soft phone to different devices The aim was to know if the port used by the VIP soft phone is static or a range of dynamic ports This was carried out using several steps Step 1 A phone call was made from the VIP soft phone on a laptop to a handset while monitoring the traffic with Wireshark The captured traffic can be seen in the figure below n capturel pcap Wireshark 1 6 2 SVN R
47. smitted and serialization are the main causes of delay and are the ones we can do something about while the other causes of delay like jitter buffers codecs are causes we can do nothing about This is where QoS can be used to prevent a voice packet from waiting on other packets before it is transmitted QoS makes the voice to be sent out of queues faster than other packets in the queue and by giving more bandwidth to voice it helps to mitigate against serialization delay In essence QoS for the voice packets reduces latency uses bandwidth well and delivers voice packets fast Using Low Latency Queuing LLQ or Priority Queuing Weighted Fair Queuing PQ WFQ commands to give priority to voice is the preferred method of configuring QoS 3 1 QoS tools for VoIP The QoS tools for VoIP are a set of mechanisms used to increase the voice quality on a network by decreasing dropped voice packets during times of network congestion and by minimizing both the fixed and variable delays encountered in voice congestion 6 The tools are divided into three categories classification queuing and network provisioning 3 1 1 Classification Classification is the process of identifying traffic into classes and grouping the identified traffic into classes It uses a traffic descriptor to categorize a packet within a specific group to define that packet 25 These traffic descriptors are based upon incoming interfaces IP precedence DSCP source or desti
48. solution Protocol request 0000 ff ff ff ff fF ff OO1f 161e f2 68 08 06 00 01 0 00 00 ins 0010 08 00 06 04 00 01 00 1f 16 le f2 68 c0 a8 f8 86 o voi h 0020 00 00 00 00 00 00 cd a8 f8 35 00 00 00 00 00 00 ww Seeth 0030 00 00 00 00 00 00 00 00 00000000 uassesse sess O Fe C Documents and Settings Ehis My Docum Packets 1249 Displayed 1249 Marked 0 Load time 0 00 406 Profi Defaut Figure 4 1 Captured Phone call Traffic to Handset From Figure 4 1 it can be seen that different protocols are being used for this single communication The TCP TLSv1 and UDP protocols are the protocols used for this particular communication from a VIP soft phone to the handset In order for the communication to be really understood a deeper analysis is required This can be seen from the different figures below TURKU UNIVERSITY OF APPLIED SCIENCES BACHELOR S THESIS Donald Egbenyon 18 DER T capture1 pcap Wireshark 1 6 2 SVN Rev 38931 from trunk 1 6 File Edit View Go Capture Analyze Statistics Telephony Tools Internals Help SAN EHXZ2S3 6 9T2 5EBIRALROIWMXI E a Filter Expression Protocol TLSv1 TCP TCP No Source Destination 62 71 106 251 62 71 106 Time 34 5 544454 35 5 544548 Length Info 1434 Application Data 1434 TCP segment of a reassembled PDU 192 168 248 121 192 168 248 121 3 51 192 168 248 121 41 5 695061 192 168 2
49. t phone to the server 62 71 106 251 an acknowledgement is sent using TCP protocol from the client computer 192 168 248 121 to the server TURKU UNIVERSITY OF APPLIED SCIENCES BACHELOR S THESIS Donald Egbenyon 19 n capture1 pcap Wireshark 1 6 2 SVN Rev 38931 from trunk 1 6 File Edit View Go Capture Analyze Statistics Telephony Tools Internals Help YSK SAXSS CeoaF LFF QQqn weUBx B Filter Expression C No Time Source Destination Protocol Length Info 34 5 544454 62 71 106 251 192 168 248 121 TLSV1 1434 Application Data 35 5 544548 62 71 106 251 192 168 248 121 TCP 1434 TCP segment of a reassembled PDU 37 5 544593 TLSVI 728 Appli ata Appl 81 Ack 192 1 41 5 695061 62 71 106 2 121 42 5 893563 i d f tree Tor br STP T t 8 45 6 413645 62 71 106 251 192 168 248 121 TCP 60 sip tls gt 53381 ACK Seg 3435 Ack 1093 46 6 413675 62 71 106 251 192 168 248 121 TLSV1 128 Application Data Application Data 47 6 433201 192 168 248 121 192 168 248 55 ICMP 74 Echo ping request id 0x0001 seg 150 48 6 439048 192 168 248 55 192 168 248 121 ICMP 74 Echo vina reolv 1d 0x0001 seq 150 Y gt Checksum 0x6679 validation disabled N 3 SEQ ACK analysis a This is an ACK to the segment in frame 37 The RTT to ACK the segment was 0 105212000 seconds f Bytes in flight 1018 Secure Sockets Layer 8 TLSV1 Record
50. the access list extended VOIP RTP and VOIP CONTROL has some packets that match it Moreover from the output below it can be seen that the policy map is being matched by the VoIP packets TURKU UNIVERSITY OF APPLIED SCIENCES BACHELOR S THESIS Donald Egbenyon 39 Figure 6 2 Policy for an interface From the previous tables and the output captured in Figure 6 4 it has been established that the required packet that needs priority is being given priority in the network Since this is the case then when QoS is configured in a LAN it is of utmost importance to VoIP traffic or any traffic that needs expedited forwarding All that is required in such a scenario is for a set bandwidth to be allocated to the VoIP traffic or other important traffic and for the other traffic that do not require priority should be given a fair share of the bandwidth Additionally networking devices usually drop packets that exceed the configured bandwidths so for voice packets to always be sent it is required that after the initial bandwidth is given to the VoIP class the network should be monitored for some time in TURKU UNIVERSITY OF APPLIED SCIENCES BACHELOR S THESIS Donald Egbenyon 40 order for the network engineer to ascertain if the bandwidth will have to be increased or not and if the configurations in such devices will be modified to prevent the dropping of VoIP packets This will help the network to deliver all traffic regardless of it bein
51. ttorin salasana vahvistus S Securi Update m J me je Jo J O p lle pois p lt J J J J O O p lle pois p lt O p lle pois p lt Op lle Opois p lt O Op lle pois p lt O Op lle pois p lt Op lle Opois p lt O O p lle pois p lt O p lle Opois p lt o o TURKU UNIVERSITY OF APPLIED SCIENCES BACHELOR S THESIS Donald Egbenyon 47 Configuration for Catalyst switch 2950 Class Maps class map match all DOM VOIP RTP description DOM VOIP RTP TRAFFIC match access group name VOIP RTP Policy Map policy map DOM QOS TRAFFIC description DOM QOS TRAFFIC POLICY class DOM VOIP RTP set ip dscp ef COS setting mls qos map cos dscp 0 8 16 26 34 46 48 56 Attaching Policy to an interface int gi0 1 service policy input DOM OOS TRAFFIC Voice RTP access list ip access list extended VOIP RTP permit ip any any dscp ef permit ip any any dscp cs5 Configuration for Catalyst switch 2960 Class Maps class map match all DOM VOIP RTP description DOM VOIP RTP TRAFFIC match access group name VOIP RTP class map match all DOM VOIP CONTROL description DOM VOIP CONTROL TRAFFIC match access group name VOIP CONTROL Policy Map policy map DOM OOS TRAFFIC description DOM QOS TRAFFIC POLICY class DOM VOIP RTP set ip dscp ef police 1000000 65536 exceed
52. y command 3 Table 5 3 Syntax Description policy policy map name Optional Display the specified policy map name class class map name Optional Display QoS policy actions for a individual class begin Optional Display begins with the line that matches the expression exclude Optional Display excludes lines that match the expression include Optional Display includes lines that match the specified expression expression Expression in the output to use as a reference point TURKU UNIVERSITY OF APPLIED SCIENCES BACHELOR S THESIS Donald Egbenyon 33 Note Though visible in the command line help string the control plane and interface keywords are not supported and the statistics shown in the display should be ignored One problem experienced in a C4506 switch is that the bandwidth command was able to be used when writing the configuration but generated several errors when the configuration was attached to an interface Why will a command be allowed that is not supported This is a question only Cisco can answer It might be purely for business purposes because even the updates released did not address these major limitations TURKU UNIVERSITY OF APPLIED SCIENCES BACHELOR S THESIS Donald Egbenyon 34 6 Results and Conclusion 6 1 QoS implementation It is necessary to carry out a QoS baseline test in order to compare the r
Download Pdf Manuals
Related Search