Multi-Protocol Label Switching Simulation - Lab Report Example

MPLS SIMULATION Introduction: MPLS, Multi Protocol Label Switching provides a great opportunity for big organizations which delve into telecommunication & data networks. When we talk about telecommunication & data networks, we consider how the data from one place to another will move without interruption. When we attend a telephone call, we receive and send sound signals to each other. Similarly when we attach our computer to the telephone line, not only we can transfer the sound signal but also we can transfer the written information like file transfer, through the modem (modulator & demodulator). Modem converts analog signal to digital signals & digital signal to analog signal. Today the internet seems to be an indispensable communication instrument, and everybody who is using the internet wants good services from internet service providers. Good service means, good downloading speed (means time related to opening, closing and downloading a file on the internet), good connectivity (Internet connection can be done easily on request), & transfer of data or file without interruption. Transfer of data or file without disturbance is a very critical factor on the internet, there are lots of internet connections providers, a lot of websites, a lot of data transfer, a lot of web server and lots of computer are accessing internet at the same time. Due to these usages, there is congestion on the internet, which results in delay in data transfer and also there is no surety that the data will be transferred completely. Today each and every organization wants to fulfill all the customers' needs. And the main agenda of most of the organization is Customer Satisfaction. Customer satisfaction is the key of success for any organization. Customer satisfaction can be achieved by providing good Quality of Service (QoS). Parameters for Quality of Service may vary from one organization to another. Here we are discussing about the organizations involve in telecommunication & data networks; the parameter of quality for this type of organization are delay, jitter, packet loss and bandwidth. So if we look into such parameters, all these parameters are very important in data transfer. Quality parameters like delay, jitter, packet loss and bandwidth are very crucial in networking. The performance of networking has complete dependence on the above four parameters. So if an organization wants to achieve customer satisfaction, they should have control on these parameters. For controlling these parameters, organizations use packet switching technology and Traffic Engineering. This technology depends on internet protocol addressing. Internet protocol addressing provides a unique number to a particular location. This unique numbers helps to find the location for transfer and minimize congestion. The internet protocol address is a twelve digit number. The Internet Engineering task force developed a technique known as Multi Protocol Label Switching (MPLS), to avoid congestion in networking, avoid delay in data transmission and keeping control on bandwidth. Multi Protocol Label Switching is based on internet protocol addressing. It involves packet data transfer. As the packet enters into Multi Protocol label switching, it receives a label. Depending upon the label Multi Protocol Label Switching defines the most suitable routing or path for data transfer. While defining the path Multi Protocol Label Switching analyze the load on the network & the type of traffic. Based on the analysis Multi Protocol Label Switching divides the traffic in a manner to minimize the network congestion. It also finds the shortest path for data transmission, which transmits the data from one location to other in the minimum time. Validation of MPLS Simulation: As we have come to know MPLS simulation is used to increase the efficiency of the network through minimizing congestion, now we have to validate that really this MPLS simulation is an effective tool for Internet service provider. There are various ways of proving or validating the MPLS simulation. For validation of MPLS simulation network simulator 2 is used. This is generally called as ns2. NS2 simulator is also known as event driven simulator. It is used in networking research. This simulator 'begun development in 1989 as a variant of the REAL network simulator' (en.wikipedia.org) and there are still some modifications and developments being done on this simulator. NS-2 is coded in C++ and TCL. The critical developments are done by Xerox PARC, UCB, and USC/ISI. Multi Protocol Label switching module is contributed for network simulator 2. This module is also known as MPLS Network Simulator. According to this model label distribution & MPLS forwarding is done by using and by LDP and CR-LDP. The main feature of this model is the capability to define and setting of a different path which is used for backing up a path. Due to this capability the MPLS can stimulate network recovery mechanism. Installation of Network Simulator 2 Before installing the Network Simulator 2, we should know the requirement of main computer that can run the network simulator. The main computer or server needs software, either windows Version using cygwin software or Linux Operating system. These systems support the installation of network simulator2. Before installing Network Simulator we should first install cygwin software. There are few steps to down load cygwin on the server: 1) Click on the internet explorer and open the website www.cygwin.com. 2) Down Load the latest version of cygwin software. 3) Click on Setup File and click the option install from internet. 4) Click on the next button to setup the root directory. 5) Click on the next button until internet connection wizard appears. 6) Select direct connection from internet connection wizard and click next button. 7) Select a website to enhance the capability of ns2, for downloading further information. 8) Click on the next. And the screen showing the progress of downloading will be appearing. Now the setup of cygwin software for supporting network simulator 2 is completed. After completion Network simulation 2 installation can be possible. For installing NS2 first check the setting of cygwin and open the cygwin software to ensure that the software is performing as per the requirement or not. After getting confirmation click on internet explorer and find www.isi.edu. Brows the site and find the latest version of network simulator 2 and down load it. While downloading store the file a specific location so that when you need that file, it can be easily found on the system. After downloading the network simulator 2 from the website open the cygwin software and install the network simulator 2. The installation of network simulator 2 will take some time but it will be automatically installed, you only need to click next button as & when required. Testing To test the MPLS Model, we should arrange the network using the following topology. Figure 1: Topology diagram This topology consists of the following: The network core is composed of 8 MPLS LSRs. These are connected with 1 mb/s links. These links are having low bandwidth which helps in its congestion. Six edge LSRs in the core access points are responsible for providing a traffic maps. 1 mb/s links is used for connecting edges with core LSR. By this way, when one packet is moving through edge LSRs, it is identified with a DSCP and kept in a category of FEC and then passed to the network core. While performing the test it should be remembered that all the LSPs used should have equal priority. Under demand mode the distribution should be made. Only six customer networks are connected to the ISP, which are having different traffic resources, such as File Transfer Protocol, CBR and Hyper Text Transfer Protocol. 2mb/s links is used to connect customer network to the border routers and border routers to the MPLS domain edge LSRS's. Flow: The voice flows are arranged in such a way that they can simulate Voice over IP flows using CODEC G.729A for compression. Each flow contains around seven simultaneous conversations, with every packet having equivalent to 30ms of conversation. The HTTP and FTP can flow the packet which is having note more that 1500 bytes. Average size of randomly chosen internet pages stimulates the HTTP flows page size. The simulated flows and its classes can be seen from below mentioned table. Flow Type Origin Destination TX Rate CIR CBS Class VOZ1 Exponential/UDP Client2 Client6 132Kb/s 150Kb/s 0Kb Gold VOZ2 Exponential/UDP Client6 Client2 132Kb/s 150Kb/s 0Kb Gold FTP1 FTP/TCP Client3 Client4 -- 400Kb/s 1Kb Silver FTP2 FTP/TCP Client7 Client5 -- 400Kb/s 1Kb Silver HTTP1 HTTP/TCP Client1 Client8 -- 100Kb/s 0Kb BE HTTP2 HTTP/TCP Client1 Client9 -- 350Kb/s 1Kb Bronze HTTP3 HTTP/TCP Client1 Client10 -- 350Kb/s 1Kb Bronze In all types of flow voice flow is given higher priority. That's why voice flows are kept in gold class. In silver class FTP flows are mapped, that's why it is compulsory that they have a higher Quality of service, at some time they can also send some traffic burst. HTTP flows have comparatively less priority, so they are mapped in last classes (bronze and best effort classes). In cases of congestion, the excess packet is dropped first. This test results proved that MPLS infrastructure has improved with regards to provision and separation when it is used jointly with different mechanisms. Results In order to confirm the results, the result of one simulation should be compared with the other simulation outcome. Here we have taken two simulations one is MPLS only and another one is MPLS with diffserv. Both the simulations were executed many times and the average result of all the attempts was recorded in table form. Standard deviation for some cases was also recorded in the Table. The following table indicates the details of total packets sent in simulation process and total number of packet rejected during each simulation. MPLS Standard MPLS+Diffserv Standard (Average) Deviation Average Deviation Total sent packets 9.146,4 50 9.244,4 11, 5 Discarded packets 19 3, 2 6, 2 1, 3 As we have already mentioned the above table is showing the comparison of two simulation process, with respect to the packet sent and discarded packets. Now we will compare the quality of service (QoS) parameters of two simulation process. The Quality of service parameters are delay, Jitter, Packet Loss & Bandwidth. All the parameters are very crucial for selecting the Multi Protocol Label Switching. Delay In all Quality of Service (QoS) parameters delay plays a significant role in voice flow metric for VoIP flows. If we analyze the table below, we observe that in MPLS model flow VOZ1 has an average delay of 206ms & flow VOZ2 has an average delay of 34ms. There was a great variation in both the values. This indicates that there is no surety of delay requirement of voice flow in MPLS model. In the MPLS with Diffserv model, the vales of VOZ1 (55ms) and VOZ2 (28ms) are very close to each other and meets with the requirements of ITU-TG.114. MPLS Standard MPLS+Diffserv Standard (Average) Deviation Average Deviation VOZ1 0,206s 0,009s 0,055s 0,0003s VOZ2 0,034s 0,0004s 0,028s 0,000005s Jitter Jitter values are also extremely important in evaluation of MPLS simulation. Jitter is also known as a pulses variation in digital transmission. Jitter can be identified by variations in signal amplitude and signal strength. Increase in jitter values may lead to interrupted transmission. The reason of Jitter includes timeouts of connections, connection time lags, congestion in data traffic, and interference. This is an undesirable output of system flaws and interruptions. To get the best knowledge of jitter, one must remember that data (whether audio, video, pictures or text) is seldom sent out wholly. Each type of data is 'broken up into small units, or packets' (Microsoft Encarta, 2005) with identifiers that indicate the sequence of the data packets when the users computer's turn to organize them for playback. When a jitter occurs, some data packets may be lost in transmission. Due to jitters, user computer may perform malfunctioning by the ways of: damaged files, corrupted audio files, interrupted phone calls on internet, time lagging or get disconnected. If we analyze the table below, we observed that in MPLS model flow VOZ1 has an average jitter of 3571s & flow VOZ2 has an average jitter of 3615s. There was a great variation in both the values. This indicates that there is no guarantee of jitter requirement of voice flow in MPLS model. In the MPLS with Diffserv model, the vales of VOZ1 (3567s) and VOZ2 (3572s) were extremely close to each other and meets with the requirements of ITU-TG.114. MPLS Standard MPLS+Diffserv Standard (Average) Deviation Average Deviation VOZ1 0,03571s 0,006s 0,03567s 0,005s VOZ2 0,03615s 0,015s 0,03572s 0,007s Figure 2: Jitter Graph for VOZ1 Flow Packet loss Packet loss is as well an imperative parameter to consider for comparing the simulation process. During simulation process packet loss should always be nil, because if packet loss is there then the actual data divided in to the packet are also lost. Due to this the file or the data may be corrupted and there is some problem in opening the file on the user's computer. Here we are comparing two simulation processes for the packet loss. The results of analysis are given in the following table: MPLS Standard MPLS+Diffserv Standard (Average) Deviation Average Deviation VOZ1 8, 6 6, 6 0 0 VOZ2 4, 2 9, 4 0 0 FTP2 2, 4 1, 1 0 0 HTTP1 0 0 6, 2 1, 3 HTTP2 0, 4 0, 9 0 0 HTTP3 0, 2 0, 4 0 0 The above table is self explanatory about the type of simulation which is more suitable for the telecommunication & data network companies. In MPLS model both voice flow, flows FTP2, HTTP2 and HTTP3 are found to have packet loss. In the MPLS + Diffserv model no packet loss has been found. And most of the voice and file transfer packets are sent. Bandwidth Bandwidth, the amount of information that can be sent through a connection between two computers in a given amount of time (Microsoft Encarta, 2005), is also an essential Quality of service parameter in simulation selection. If we analyze the below table, we found that in all flows the bandwidth is increasing from MPLS model to MPLS with diffserv model except in HTTP1. If we see File transfer protocol1 (FTP1), It has an increase 25698Kbit/s to 29051Kbit/s. This indicates that the MPLS with Diffserv model also is capable of providing quality of service of the bandwidth. MPLS Standard MPLS+Diffserv Standard (Average) Deviation Average Deviation VOZ1 133,52Kbit/s -- 139,21Kbit/s -- VOZ2 131,39Kbit/s -- 141,31Kbit/s -- FTP1 256,98Kbit/s -- 290,51Kbit/s -- FTP2 215,81Kbit/s -- 229,37Kbit/s -- HTTP1 47,65Kbit/s -- 24,74Kbit/s -- HTTP2 40,48Kbit/s -- 43,06Kbit/s -- HTTP3 38,53Kbit/s -- 42,81Kbit/s -- Figure 3: Bandwidth Received for Flow FTP1 Figure 4: Bandwidth Received for Flow HTTP1 References: Wikipedia, the Free Encyclopedia. 24 June 2008. Wikipedia. 11 July 2008 . Harrison, Peter John. "Bandwidth." Microsoft Encarta 2006 [CD]. Redmond, WA: Microsoft Corporation, 2005. "Packet." Microsoft Encarta 2006 [CD]. Redmond, WA: Microsoft Corporation, 2005. Read More