Introduction To QoS For Network Management Essay.

What is QoS and how it helps manage congestion?

Quality of Services (QoS) is need to be introduced to the network of the company for the management of congestion with a discussion of service classes and architecture. QoS mechanisms are described that helps with congestion (Albur et al. 2020). QoS helps in prioritizing traffics for controlling jitters and delays during congestion. Further, the QoS policy will be devised with the help of a QoS mechanism. Narration is provided on traffic parameters and the QoS tool.  Two scenarios are discussed using the policies with appropriate considerations. Lastly, a QoE strategy is devised with its Planning, lab-testing, monitoring, and mapping stages.

The firm is looking to adopt network management to address real-time issues of the network. The employees in the branch are connected to the headquarters using a leased-line connection offering 2Mbps bandwidth allowing them to browse the web and communicate via Voice-over-IP calls using a company-built application. Thus, QoS Management is introduced to manage congestion in the low bandwidth system. Quality of services (QoS) is termed as a set of vital technologies running on a network that helps in facilitating critical applications to operate and allow traffic under the limited capacity of the current network. QoS management helps in supporting the prioritization of applications important to the system and reducing overall congestion. QoS helps in tracking the traffic on the network, identifying risks of network congestions to guide the vital tasks to a smoother course of traffic (Azamuddin et al. 2020). Congestion management helps in identifying the marking of the packets and facilitates queuing the marked packets based on the queuing system.    

The most common service classes found in QoS management are differentiated services, predicted services, and Guaranteed services. Differentiated services are used for allowing the networks for direct configuration of relevant parameters over the routers and switches rather than using the resource reservation protocol (Biyase 2021).  It helps in assigning policies to the specific interface for controlling traffic. In predictive classes, the network allows good bandwidth on average. The architecture helps in the specification of the elements provided for guaranteeing the quality of services. The Flow specification consisting of the Rspec and Tspec is specified accordingly by the client. Admission control helps in managing Rspec and Tspec to handle the network without harming the current flow of service (Osso 2018). Resource reservation is dome by reservation protocol reserving the flow of the packets. Lastly, the packets are classified accordingly with the reservation of flow with the help of Packet scheduling. 

The basic mechanism that should be followed is the development of understanding on Packet scheduling, management of congested interfaces, the understanding of the network on which packet to be delivered first when multiple packets are waiting to be delivered and the knowledge on how the experience of the user can bear with packet jitter, loss, and latency. Packet scheduling can help with the proper allocation of resources in a shared network among the various flows in the network for maximization of the utilization and targeted sharing of flow (Proskochylo, Zriakhov and Akulynichev 2018). The networks should transmit the proper amount of data accordingly to the threshold of the network. The flow of packets in the network is needed to be regulated so that no packet has to wait for delivery and the decision done by the network device should be correct to erase congestion (Sachdeva and Dev 2021).  Lastly, it should be determined how much tolerance does the client has for jitters, loss, and latency in the network.

Different QoS mechanisms and service classes

Discussion of traffic parameters that are needed to be prioritized for VoIP conversation:

The company has a bandwidth of 2048kbps which helps the customers to conduct VoIP calls and web browsing. According to the class-based weighted fair queueing bandwidths are allocated according to the weights designated to the traffic classes. The ratio in the Traffic class AF41, AF31, AF21 is 2:1:1 and the remaining part is unallocated (Pesántez-Romero et al. 2022). So, the ratio between traffic class and unallocated data will be 512:256:256:1024. The scheduler present helps in sending packets according to the clock cycle. This mechanism is improved by low latency queueing by adding priority queueing to class-based weighted fair queueing for avoiding delays and mitigating jitters. The parameters to be followed are that the VoIP traffic should be delivered as loss of delivery might lead to an audible drop in the conversation taking place. The VoIP packets must be sent through the VoIP traffic on time to minimize delay and jitters resulting in stutters in audio (Issing 2018). There should be no prioritization of traffic class of different bandwidths as the policy is set for balancing timely delivery of VoIP packets taking care of other classes required to use the network. The traffic class is set to be flexible and provide easy configuration.

Tools like Weighted Fair Queueing and class-based Weighted Fair Queueing cannot provide the guarantee for achieving low delay and secured bandwidth to applications of VoIP due to the lack of priority queues. Application of VoIP requires to have minimum delay and mitigated jitters or delay variations. The Low Latency Queueing used for this network over other available queues as it helps in VoIP call satisfaction being a strict-priority queue helps in facilitating guarantee for minimum bandwidth, and rigidity in not allowing to transit data over permission during times of congestion. The traffic arriving over the limits of the queue is dropped. Thus, starvation can be mitigated in the queues during the congestion time, as there is available bandwidth for transmission even in traffic. Low latency queueing has all the properties of CBWFQ giving guarantee of a defined bandwidth for the classes. The LLQ uses priority queues to handle the delay constraints occurring in real-time (Saboka 2021). CBWFQ benefits in ensuring acceptable throughput to the traffic classes with less delay sensitivity. Sensitive data are sent to priority queue and the priority queue gets emptied before any other queue is served. Time-sensitive scenarios are given priority minimizing delays to execute the work. LLQ is common in VoIP maintaining traffic in network.        

The scenario that is taken in the first place was a QoS policy shaped for facilitating all the individual employees of the branch simultaneous conversation over the VoIP calls with the guarantee of a low fidelity mode with no service interruption. The QoS mechanism that was relied on will be use of low latency queueing. The limit of capacity of the QoS policy is 800Kbps. The fidelity modes provided by the company namely High and Low, help them to communicate without any audio lags, delays. The corresponding bit rate can be calculated by the size of the transfer (bytes) divided by the transfer time (in second) and multiplied by eight. The company also allows its employees to use web browsing and handle other tasks over the internet websites. The LLQ mechanism will be used on this system to help prioritize the needs of the employees using the network by using the priority queue along with the benefits of CBWFQ (Bensalah, Bahnasse and El Hamzaoui 2019). The first scenario can be handled by using priority queue for the main needs of the employees helping it to be facilitated first. Thus, the network prioritizes the simultaneous calling in low-priority mode for all the employees. The help provided by CBWFQ helps in giving guarantee to bandwidth for other tasks in network like web browsing, high-fidelity calls and transactions. Users can define the number of required queues in the system. Users can define the classes helping them to understand the priorities of the tasks. The bandwidth allocation can be set on the user-definable interface of the company. The traffic classes are made by using MQC. The traffic of simultaneous calls in low fidelity mode is provided a strict-priority queue for being delay-sensitive. None of the other traffic classes in the link is starved at any point of the queuing. The queue is strictly policies by 40% of the bandwidth.         

Understanding packet scheduling and management of congested interfaces

The second scenario is that the company needed a QoS policy relying on a present QoS mechanism to facilitate 10 employees to conduct simultaneous conversation over VoIP with the provided guarantee of High-fidelity mode and without any interruption occurring in facilitating services. The mechanism to provide sense to the scenario is the Low latency queueing. This combination of strict priority queueing with the presence of all the benefits of a class-based weighted fair queueing helps to guarantee the execution of the delay-sensitive traffic first (Gupta, Panwar and Goel, 2020). The company facilitates two fidelity modes that help them to interact among themselves over the internet without delay or lag in the audio. The company network allows a low and high-fidelity mode with 32kbps and 128kbps bitrate simultaneously. Thus, the priority set for the scenario will be to provide high-fidelity calls simultaneously to 10 individual employees, along with guaranteed traffic for other activities like web browsing, information transaction and low-fidelity calls to all the members in the network. The highest priority task is sent to the highest position of the queue getting delivered in a timely manner. Network delays and packet losses are needed to be mitigated to provide reliability and accessibility. The policy states that if any packet exceeds the requested bandwidth after using the priority queue should be discarded immediately if interface is found busy. With this scenario too, users can determine the number of queues required. Users can help define classes by MQC according to the priorities and delay-sensitivity. The corresponding bitrate for the priority is 40% of the bandwidth. None of the traffic classes constructed are starved at any point and guaranteed with fixed bandwidth accordingly. The network capacity limit for the policy is around 40% of the bandwidth allocated to the high-fidelity calls for the 10 employees. Rest of the bandwidth is allocated accordingly to remaining low-level traffic classes.

The quality of experience (QoE) of an end-user is determined by the impact of the behavior of the network on them. Measurement is done by using subjective tests, with mean opinion score (MOS) metrics (Saburova, Abdulbari and Kadatskajan 2021). The subjective test planning helps in determination of the different values of QoE, starting from Excellent, Good, Fair, Poor and Bad for the test. The QoS metrical values are determined for the policy discussed. The values help in determination of the level of quality of service needed to be discussed before implementation of the policies in the network (Daengsi, Sirawongphatsara and Wuttidittachotti 2021). QoE helps in understanding the individuals according to their level of satisfaction. The different values of QoS in the considerations starts from good which are Reliability and Availability. The intermediate values determined are throughput and bandwidth of the internet. Finally, the bad values are delay, jitters and lags.

Lab testing helps in definition of the good and bad qualities of QoE in the proposed scenarios. The subjective test taken helps in choosing its application in the various domains to be understood. VoIP, web browsing and various other factors are determined which affect the quality of services like delays, lags and jitters (Bouraqia et al. 2020). The plan is derived to test the qualities of QoE, starting from a test server which is a dedicated server used for testing QoS. The test is conducted to test the effects to QoE determining level of satisfaction. The client interface is checked to determine the qualities. The campaign is set up to check the test server and client interfaces checking the values.

Allocation of bandwidth for VoIP calls

The system works as a subjective test determining the values of QoE for services of QoS. The system helps in identifying the level of QoE by judging the connections that change the QoS determined in the testing phase by use of a questionnaire. The metrics determined are evaluated by the employees according to their level of satisfaction from the experience. The monitoring stage helps in estimating the QoS factors of the operational system (Metzger et al. 2018). The bad value should correspond to the worst experience gained from the network. Poor would be a bad yet satisfactory outcome of the network metric. Fair would be a neutral or quite satisfactory outcome that can be achieved by the network. Good will be a level where expectations are met quite a lot and experience is better than fair. Lastly, excellent is the best level of performance meeting all your expectations properly. The employees in the branch are involved by giving them the right roles in the intervention. Training is provided for better understanding. The tasks are allotted according to their level of proficiency to check the system. The feedbacks are taken from the employees to keep them engaged and make them understand the various values. 

Mapping is done on the values determined from monitoring to categorize the metrics that change the system. Tables are made according to the values obtained. Mapping helps in facilitating analysis of the relations of the metrics and their values with the QoE/QoS of the company network (Hu et al. 2020). The causes are linked with its parameters by the users. Records are mapped with the results of the monitoring of the subjective tests performed. It helps in understanding the level of satisfaction of the users from the experience of the network. The values of QoE are understood how they contribute to the QoS in the company.

Conclusion:

Network management is necessary with QoS management to regulate congestion management. Understanding is gained on the QoS management being important to congestion management. Two mechanisms are described to address congestion management. QoS policy is devised relying on the QoS mechanisms with discussion of trafficking parameters needed to be prioritized for the VoIP calls and low-latency queueing as suitable tool. Two scenarios are taken along with suggested considerations for the networks. Lastly, Quality of experience strategies is made with suggestive assessment determined with planning, lab-testing, monitoring and mapping.  

References:

Albur, N., Handigol, S., Naik, S., Mulla, M.M. and Narayan, D.G., 2020, September. QoS-aware Flow Management in Software Defined Network. In 2020 12th International Conference on Computational Intelligence and Communication Networks (CICN) (pp. 215-220). IEEE.

Azamuddin, W.M.H., Hassan, R., Aman, A.H.M., Hasan, M.K. and Al-Khaleefa, A.S., 2020. Quality of service (Qos) management for local area network (LAN) using traffic policy technique to secure congestion. Computers, 9(2), p.39.

Bensalah, F., Bahnasse, A. and El Hamzaoui, M., 2019, May. Quality of service performance evaluation of next-generation network. In 2019 2nd International Conference on Computer Applications & Information Security (ICCAIS) (pp. 1-5). IEEE.

Biyase, L.Z., 2021. Scalable Bandwidth Management in Software-Defined Networks (Master's thesis, Faculty of Engineering and the Built Environment).

Bouraqia, K., Sabir, E., Sadik, M. and Ladid, L., 2020. Quality of experience for streaming services: measurements, challenges and insights. IEEE Access, 8, pp.13341-13361.

Daengsi, T., Sirawongphatsara, P. and Wuttidittachotti, P., 2021, December. Proposed QoE Models Associated with Delay and Jitter Using Subjective Approach and Applications for 4G and 5G Networks. In 2021 4th International Conference on Advanced Communication Technologies and Networking (CommNet) (pp. 1-4). IEEE.

Gupta, S., Panwar, A. and Goel, S., 2020, March. A Comparison of Voice-over-IP Capacity and Performance for Different Scheduling Techniques. In 5th International Conference on Next Generation Computing Technologies (NGCT-2019).

Hu, Z., Yan, H., Yan, T., Geng, H. and Liu, G., 2020. Evaluating QoE in VoIP networks with QoS mapping and machine learning algorithms. Neurocomputing, 386, pp.63-83.

Issing, J., 2018. Adaptation to Varying Network Conditions and Conversation Patterns for Robust High Quality Audio Communication over the Internet Protocol. Friedrich-Alexander-Universitaet Erlangen-Nuernberg (Germany).

Metzger, F., Hoßfeld, T., Skorin-Kapov, L., Haddad, Y., Liotou, E., Pocta, P., Melvin, H., Siris, V.A., Zgank, A. and Jarschel, M., 2018. Context monitoring for improved system performance and QoE. In Autonomous Control for a Reliable Internet of Services (pp. 23-48). Springer, Cham.

Osso, R., 2018. Handbook of emerging communications technologies: the next decade. CRC Press.

Pesántez-Romero, I.S., Pulla-Lojano, G.E., Guerrero-Vásquez, L.F., Coronel-González, E.J., Ordoñez-Ordoñez, J.O. and Martinez-Ledesma, J.E., 2022. Performance evaluation of hybrid queuing algorithms for QoS provision based on DiffServ architecture. In Proceedings of Sixth International Congress on Information and Communication Technology (pp. 333-345). Springer, Singapore.

Proskochylo, A., Zriakhov, M. and Akulynichev, A., 2018, October. The effects of queueing algorithms on QoS for real-time traffic in process of load balancing. In 2018 international scientific-practical conference problems of infocommunications. Science and technology (PIC S&T) (pp. 575-580). IEEE. 

Saboka, B., 2021. Improving the Quality of Service of Voice over Internet Protocol in Ethio Telecom Service Level Agreement Customers (Doctoral dissertation, St. Mary’s University).

Saburova, S.A., Abdulbari, G. and Kadatskaja, ?.I., 2021. Methods Of Control Quality Of Services VoIP Over LTE.

Sachdeva, R. and Dev, A., 2021. Review of opportunistic network: assessing past, present, and future. International Journal of Communication Systems, 34(11), p.e4860.