Telecommunications and Networking - Assignment Example

Telecommunications and Networking Contents Question 3 Reliability 3 Delay (latency) 4 Jitter 5 Bandwidth (throughput) 7 Question 2 8 Scheduling 8 Traffic Shaping 10 Admission control 11 Resource reservation 12 References 14 Question 1 Reliability Within the scope of Quality of Service (QoS), there are various elements of network performance such as reliability. In telecommunications and networking the term reliability means to ensure that the data being delivered by the source arrives at the destination end. Reliable protocol is one which encompasses reliability properties in terms of data delivery in comparison to unreliable protocol that does not send any type of notifications to sender in relation to data being transmitted. There are often more overheads in reliable protocols compared to unreliable protocols that in turn make them less scalable and slower. For multicast protocols this happens to be a big issue but has no such affect on unicast protocols. TCP which is supposedly the major protocol utilized in Internet can be stated as reliable unicast protocol. UDP on the other hand can be defined as unreliable protocol used in computer games and other Internet applications where speed is a major factor of concern and loss of data is not significant due to nature of data being transitory. Reliability measures can be divided into many factors such as source node communicating with end or terminal node for every origin destination or OD pairs and this is known as terminal reliability (Kurose and Ross, 2007). Reliability of many source to terminal node can be stated as (k>1) nodes where k is the terminal communicating with terminal node. Reliability in terms of source to many terminals is source interacting with “k” operating terminal node. Network reliability is communication established between all operating nodes. Figure 1: Network Reliability (Source: Kurose and Ross, 2007) As per figure 1 there are several nodes between s (source) and t (terminal) and hence reliability measures can further include node to node blocking, functional reliability, node to node delay and reliability of average terminal. The major constraints for reliability are performance and cost. However in order to set high reliability it is essential that reliability of individual components is increased across the network, intermediate links between destination and origin nodes are decreased and alternative paths are increased that are available to OD pairs. Delay (latency) Latency or delay is another vital element that is related to QoS and determines network speed. It basically refers to various kinds of delays that are incurred during network data processing. A network connection can be stated as possessing low latency when it experiences less of delay times. On the contrary a high latency network is one that is subjected to long delays while data is being transmitted from source to destination. This factor greatly impacts the quality of service in telecommunications and networking. Delay majorly refers to the travelling time taken by a bit of data to reach from one node to another. The measurement of such network delay or latency is taken in fractions of seconds. Delay amongst networking may differ on the basis of specific pair location of the communicating nodes. Precise measurements are undertaken to clearly judge the factor or factors which are leading to such overall network delay. The delay or latency factor is categorized into four major parts such as processing delay, queuing delay, transmission delay and propagation delay. Processing delay is the excess time taken by time routers in order to process overall packet header. Queuing delay is the time spent by packet header in relation to routing queues. Transmission delay can be classified as time taken to push bits of packet onto the link. Propagation delay is the overall excess time encountered by a signal so as to reach its destination. There exists a minimum delay level that is experienced due to time taken for serial data packet transmission through a specific link. Some variable delay level is added to the existing delay because of network congestion (Kitawaki and Itoh, 2005). Network delays related to IP can vary from few milliseconds to hundred milliseconds. This factor is usually noticed to be low in case of file transfer, emailing, audio or video streaming but it is relatively high in video conferencing or telephony majorly due to propagation delay. Jitter Jitter in relation to clock source is a deviation of periodic signal from true periodicity in telecommunications and electronics. Jitter is usually observed in characteristics such as signal amplitude, successive pulse frequency and periodic signals phase. In design of communication links jitter can be considered to be an undesirable but significant factor. TCP/ IP networks have certain inherent tendencies that result into such undesirable effects. Figure 2 displays the composition of data packets due to jitter. Figure 2: Effects of Jitter (Source: Iannone, 2011) As shown in figure 2 the sender transmits data in a continuous fashion and even spaces them at equal distance but due to configuration errors, network congestion or improper queuing there arises a delay in data packet transmission which disrupts their constancy. This delay causes a major problem in applications such as audio streaming, video streaming, telephony or video conferencing as receiver needs to wait for variable data packets that are arriving late and this eventually causes an impact on quality of service of such applications. Jitter can even be quantified in peak to peak displacement or root mean square like any type of varying signals. It can even be illustrated in spectral density as any other varying signal. Jitter can even be caused by crosstalk with signal’s carriers or by EMI (Electromagnetic Interference). In networking jitter can be stated as latency variation as measured in terms of time variability of data packets across the network. A network will have no jitter if it possesses constant latency. Packet jitter or packet delay variation is the average deviation from mean latency of the network. Packet delay variation or PDV can be regarded as an important factor of QoS or quality of service in terms of network performance (Comer, 2008). Anti-jitter circuits and jitter buffers are majorly incorporated to establish QoS in telecommunications and networking. Bandwidth (throughput) In networking bandwidth can be defined as a measurement of available bit rate or consumed resources of data communication expressed in bits in multiples or per second. Bandwidth in computer networking is associated with data rate that is supported by an interface or computer network. It cannot be stated as the only factor that has contribution towards the speed of a network. Latency plays an active role in determining the performance of a computer network. Bandwidth (throughput) is considered to be a primary measure of speed of a computer network. In present scenario Internet applications as well as network products that are been sold in the market place has their bandwidth clearly mentioned for knowledge of customers. This factor basically represents the overall capacity that is possessed by a network connection. The capacity being large signifies that better performance is exhibited through the network. Bandwidth can be associated with theoretical as well as actual throughput. Like for instance the peak bandwidth that is supported by dial up modem is 56 kbps but because of physical limitations of various factors such as telephone lines, etc., the actual bandwidth that is supported by modem is 53 kbps. On the other hand Ethernet networks can actually support till 1000 Mbps which is the maximum bandwidth but this cannot be achieved majorly because of overhead in operating systems and computer hardware. High bandwidth network is associated with high speed that is generated by the network system. Quality of service that is linked with such internet applications depends greatly on their speed. Video streaming and video conferencing generally exhibit high bandwidth whereas email and telephony are associated with low bandwidth. There are various tools that are employed so as to measure the bandwidth of a network. In LAN’s the tools used are ttcp and netperf. Internet offer various online bandwidth testing programs for customers (Forouzan, 2007). However even these tools are not useful as utilization of bandwidth is not easy to measure since it varies with time. This factor depends on the hardware configuration and even characteristics and usage of software applications. Question 2 Scheduling The increasing need to meet or over exceed customer’s expectations poses a high requirement to improve upon the QoS or quality of service. There are various applications designed for this improvement and one amongst them is scheduling. This technique enables a packet to indulge into link transmission amongst various packets that are being stored in a specific buffer or multiplexing point. The scheduling algorithms majorly focus on selecting that packet which can satisfy the needs of QoS. There are various components of such algorithms firstly it operates across those points that are multiplexing by nature. It can be easily implemented in high speed hardware. The scheduling algorithm helps in regulating interactions across flows and is highly dependent on the techniques associated with buffer management. This technique states that in order to simplify a problem and make it independent by nature it is essential that infinite number of capacity buffers is taken into consideration. Figure 3 given below clearly describes the way in which scheduler helps in enhancing QoS across the network route. Figure 3: QoS-capable router (Source: Stiliadis, 2007) As per figure3 scheduler in the technique plays an active role to choose appropriate packet that needs to be transmitted across the network to the output link. The properties of the scheduling algorithm are that it facilitates flow isolation and deterministic guarantees or end to end statistical. Flow isolation indicates that non conformant flows should not cause any impact on the conformant flows. In this context queuing per flow indicates that scheduler is the one who would select the queue from which data packet needs to be transmitted. The other property signifies the rate of delay or bit rate that is specific for a particular flow or equal for every data flow. Scheduling algorithms for QoS can be classified into two schedulers such as work conserving scheduler and non-work conserving scheduler (Stiliadis, 2007). In work conserving scheduler optimal performance is achieved in terms of throughput and there is continuous packet transmission till there is availability of one packet in switch buffer. On the contrary in other scheduler there is improvement in terms of buffer size reduction. Traffic Shaping Traffic shaping can be considered to be rate limiting and is a well known technique of computer network traffic management. This procedure delays all or some of the datagrams so as to bring into compliance with specific traffic profile. It is generally incorporated so as to improve latency, optimize performance and increase in usable bandwidth for certain data packets whereas delaying the rest. This technique apart from enhancing QoS also serves as an effective tool for government in terms of restricting access to particular services or delaying packets that consists of information in relation to censorship. However in such case users often experience service that is of low quality and receives an impression that is misleading in terms of a site that is unreliable or slow and this may eventually lead to other site’s preference amongst users. In peer to peer networks of file sharing ISPs at times utilize traffic shaping like BitTorrent. Latency usually arises when a link gets saturated to the extent of significant contention level. Traffic shaping helps to eliminate this issue from the system and continuously monitor latency in computer network. This technique provides a control means for traffic volume being sent across a network in specific bandwidth or rate limiting that is traffic being sent at maximum rate. However this method of control can be achieved through various means but traffic shaping can be majorly accomplished through delaying data packets (Evans and Filsfils, 2007). It is usually applied at the edges of network in order to control the overall traffic that is entering into network but can even be applied by any network element or traffic source. Metering is employed in traffic shaping and the packets identified through the procedure are stored in FIFO buffer. There are four basic steps of implementation process such as identifying overflow condition, traffic classification, self limiting sources and then forming relationship with traffic management. Admission control Admission control can be considered to be a validation process that is been utilized in networking and telecommunications. It is form of a check that is executed before any establishment of connection in order to verify whether for proposed connection current resources would be sufficient or not. Admission control methods usually give rise to access control methods that helps to provide quality of service. In order to ensure that quality is maintained throughout the network there is a need of control access of flow of information. Admission control methods provide this form of control in regards to networking. The major goal of these methods is to identify the expected bandwidth for the data stream that is incoming and utilize all of the resources in order to prevent congestion (Yerima, 2011). These methods are utilized for jitter sensitive or delay sensitive services or rather for all those real time applications. Some of the admission control techniques are based on statistical indicators or mathematical calculations while others are related to measuring traffic. However the broad categories for these two methods are MBAC which stands for measurement based admission control and PBAC or parameter based admission control. PBAC method is related to present active traffic and characteristics of active traffic in total. This particular method cannot be stated as optimal as it is not dependent on any new traffic that is incoming in the network. On the other hand MBAC method is not about obtaining source’s parameters but to make measurement of the real time running of network. It achieves high utilization of network because it serves new incoming traffic. However the common objective of these methods is to possess the ability of allocating bandwidth for flow of entire set of actual data in scenario where there is no option of exceeding the line’s total capacity (Alipour and Mohammadi, 2008). There needs to be a direct association of QoS parameters and the respective nodes on which such method is been applied. Resource reservation Resource reservation is another method that has been introduced so as to enhance the factor of QoS in telecommunications and networking. The protocol of resource reservation can be regarded as transport layer protocol that is framed so as to reserve resources throughout the network. This protocol facilitates resource reservation through a setup that is receiver initiated both for unicast as well as multi-cast flow of data with robustness and scaling. However this method is not engaged into any form of application data transport but works in similar fashion as that of control protocol. This method can be utilized by routers or hosts so as to provide QoS for data stream application. The protocol has certain attributes such as demand for simplex flows in which traffic flows in one direction from sender to multiple receivers. It is receiver oriented protocol in which resource reservation and initiation of data flow is maintained at the receiver end. There are two types of messages on basis of which resource reservation technique is applied that are path messages and reservation messages or resv. The path message encompasses the IP address and is transmitted from the sender node across the data path. Resv message is sent across the reverse data path from receiver to sender. This message even contains flowspec data in order to identify all those resources that are needed by the flow. The entire operation of resource reservation can be classified into three steps firstly is making reservation entirely based on request parameters. In this step either the packet classifier is included to handle data packets or negotiation is done with upper layers in terms of performance of packet handling (Gaiti, 2007). The next step is forwarding the request in upstream direction where the resv message can be easily modified by the forwarding node. Finally the routers store nature of information flow and police it so that the issue of reservation being cancelled due to crash of sender and receiver can be resolved. References Alipour, E., and Mohammadi, K., 2008. Adaptive admission control for QoS guarantee in differentiated services networks. International Journal of Computer Science and Network Security, 8 (6). pp. 93–98. Comer, D. E., 2008. Computer Networks and Internets. New Jersey: Prentice Hall. Evans, J., and Filsfils, C., 2007. Deploying IP and MPLS QoS for Multiservice Networks: Theory and Practice. Massachusetts: Morgan Kaufmann. Forouzan, B. A., 2007. Data communications and networking. New York: McGraw-Hill. Gaiti, D., 2007. Network Control and Engineering for QoS, Security and Mobility. New York: Springer Science & Business Media. Iannone, E., 2011. Telecommunication Networks. USA: CRC Press. Kitawaki, N., and Itoh, K., 2005. Pure Delay Effects in Speech Quality in Telecommunications. IEEE, 12(5), pp. 435-438. Kurose, J. F., and Ross, K. W., 2007. Computer Networking: A Top-Down Approach. London: London Publishing. Stiliadis, D., 2007. Hardware Implementations of Fair Queueing Algorithms for Asynchronous Transfer Mode Networks. IEEE Communications Magazine, 27(1), pp. 112-115. Yerima, S., 2011. Implementation and evaluation of measurement-based admission control schemes within a converged networks QoS management framework. International Journal of Computer Networks & Communications, 3 (4), pp. 137–152. Read More