Average Queue Size Research Articles

A New Modified Dropping Function for Congested AQM Networks

Active queue management schemes are used to reduce the number of dropped packets at the routers. Random early detection uses dropping probability which is calculated based on the average queue size. Further it is modified according to the value of the count indicating the number of unmarked packets that have arrived after a marked packet. The impact of random variable i.e. number of packets arrived after a marked packet over the dropping pattern is investigated. The proposed model achieves smooth dropping pattern which results in improvement of quality of service parameters. A new model for dropping probability is proposed with different dropping function. The effect of new dropping probability results in the increase of the throughput and reduction of the expected end-to-end delay. An important finding is that the choice of modified dropping function significantly affects the performance measures of the networks.

Hybrid Fuzzy Congestion Controllers for Computer Networks Tuned by Modified Particle Swarm Optimization

One of the most debated issues nowadays is the quality of computer network service. The best internet service must provide a fast processing of the traffic. Each router has a queue of packets that provides a buffer space, where the packets wait for processing. Transmission Control Protocol (TCP) is a packets congestion control theory. Active Queue Management (AQM) is a mechanisms proposed to employ at gateways to improve the performance of TCP congestion control. AQM mechanisms aim to provide high link utilization with low loss rate and low queuing delay while reacting to load changes quickly. Random Early Detection (RED) is an extensively studied AQM algorithm that can detect congestion by dropping packets randomly with certain probability that serves as the function of the average queue size. In this work, hybrids Fuzzy Logic Controllers (FLC) are proposed to measure the router queue size directly by use them as a congestion controllers. A multiple hybrid fuzzy controllers are proposed, where (Proportional Integral Derivative controller (PID) -like FLC-Particle Swarm Optimization (PSO) Based, Proportional Derivative (PD)-like FLC with conventional I-PSO Based and PID tuned by Fuzzy Logic-PSO Based), which is provided to regulate the queue length, round trip time and packet loss. The Particle Swarmed Optimization (PSO) algorithm is used for tuning the gains of hybrid fuzzy logic controller which helps in reducing the error of the queue size. This is achieved through minimizing the rise time, peak time, settling time and overshoot of the AQM response. The empirical results revealed a high-performance improvement regarding the proposed method in comparison to previous works of other researchers.

Performance Analysis of AQM Scheme Using Factorial Design Framework

The active queue management (AQM) is a router assisted congestion control technique which improves the performance of the networks. The AQM algorithm handles the congestion by dropping the packets from a congested router where dropping probability is computed with the help of average queue size, loss-rate, queuing delay, or other parameters. A novel approach based on design of experiments is proposed to study the performance measures related to several AQM schemes viz., random early detection (RED), random exponential marking, modified RED, adaptive RED, stabilized RED, three-section RED, and AQM with random dropping. The impact of several input factors on the performance measures viz., throughput, queuing delay, loss-rate is investigated by using the factorial design where it is used to find the interaction of input factors. The relative changes in the output responses on account of changes in input factors or variables are evaluated. Sensitivity analysis is carried out by computing the weighted sum of relative changes of response variable with respect to input factors for each AQM scheme. Based on the sensitivity analysis we find that a new AQM with random dropping is most robust while in contrast three-section RED is least robust.

Bias correction for estimation of performance measures of a Markovian queue

There are several situations in our daily lives in which queues are present, such as cafeterias, supermarkets, banks, gas stations, and so forth. The performance of such queues can be described by several measures. In this article, the focus is on estimates of traffic intensity ( $$\rho$$ ), also called the utilization factor of the service station, the expected number of customers in the system (L), and the average queue size ( $${L_{q}}$$ ) for infinite single-serve queues with Poisson arrivals and exponential (Markovian) service times. The computational experiments show that the maximum likelihood estimators (MLEs) of the performance measures are biased for small and moderate samples ( $$n < 50$$ ). Thus, a version corrected by the nonparametric bootstrap method is analyzed, demonstrating that researchers could achieve with an extra computational effort bias-corrected estimates for samples of size $$n=10$$ with average errors equivalent to the estimates from the MLE for samples of size $$n=200$$ . This reduction can be very important in practical applications because of the cost and time reduction that it may bring to the process of estimating the performance measures of a queueing system.

A stochastic approximation approach to active queue management

Recently, a dynamic adaptive queue management with random dropping (AQMRD) scheme has been developed to capture the time-dependent variation of average queue size by incorporating the rate of change of average queue size as a parameter. A major issue with AQMRD is the choice of parameters. In this paper, a novel online stochastic approximation based optimization scheme is proposed to dynamically tune the parameters of AQMRD and which is also applicable for other active queue management (AQM) algorithms. Our optimization scheme significantly improves the throughput, average queue size, and loss-rate in relation to other AQM schemes.

A Late Night at the Virginian Restaurant

In this exercise, the manager of The Virginian, a small restaurant and bar in Charlottesville, has asked for help in reducing the average queue size at its bar. The Virginian prides itself on its fast and consistent service. (View the video at http://thevirginiancville.com/.) Students are offered four scenarios to guide their analyses. Students conduct analyses by preparing a simulated model on their own or using a simulation created with AnyLogic software.This exercise is supported by A Late Night at The Virginian Restaurant: Model Documentation and Guide (UVA-OM-1553). After learning the steps to run the model, students are asked to run it for four hours of simulated time under the four scenario conditions. Excerpt UVA-OM-1551 Dec. 2, 2016 A Late Night at The Virginian Restaurant The manager of The Virginian, a small restaurant and bar in Charlottesville, has approached you and asked for help in reducing the average queue size at its bar. The Virginian prides itself on its fast and consistent service. (View the video at .) The uniquely small size and location of the bar make The Virginian's service problem particularly challenging. The bar is especially narrow, which means there is only an opportunity for one long queue to form for customers ordering beer, wine, and other beverages. Right now the bar manager believes that there are, on average, around 10 customers in line at one time between 10:00 p.m. and 2:00 a.m. This makes the bar area very crowded and hurts business. The bartenders estimate that customers arrive at The Virginian randomly at an average of three per minute. The manager would like to get the average queue size below two, using the fewest number of bartenders possible. The system estimates that three new customers desire a drink every minute. Right now there are two bartenders. How many bartenders would the manager need to get the average queue size below two customers? Run the model for four hours of simulated time. 1. Scenario 1: Service time is triangular; it takes a bartender between 15 seconds and 75 seconds to fill a drink order, the most common time being 30 seconds. . . .

Probabilistic Cooperation of a Full-Duplex Relay in Random Access Networks

In this paper, we analyze the probabilistic cooperation of a full-duplex relay in a multiuser random-access network. The relay is equipped with on/off modes for the receiver and the transmitter independently. These modes are modeled as probabilities by which the receiver and the transmitter are activated. We provide analytical expressions for the performance of the relay queue, such as arrival and service rates, stability conditions, and the average queue size. We optimize the relay’s operation setup to maximize the network-wide throughput while, simultaneously, we keep the relay’s queue stable and lower the relay’s receptions and transmissions. Furthermore, we study the effect of the SINR threshold and the self-interference coefficient on the per-user and network-wide throughput. For low SINR threshold, we show under which circumstances it is beneficial to switch off the relay completely, or switch off the relay’s receiver only.

Bufferbloat Avoidance with Frame-Dropping Threshold Notification in Ring Aggregation Networks

In recent years, the reduced cost and increased capacity of memory have resulted in a growing number of buffers in switches and routers. Consequently, today's networks suffer from bufferbloat, a term that refers to excess frame buffering resulting in high latency, high jitter, and low throughput. Although ring aggregation is an efficient topology for forwarding traffic from multiple, widely deployed user nodes to a core network, a fairness scheme is needed to achieve throughput fairness and avoid bufferbloat, because frames are forwarded along ring nodes. N Rate N+1 Color Marking (NRN+1CM) was proposed to achieve per-flow fairness in ring aggregation networks. The key idea of NRN+1CM is to assign a color that indicates the dropping priority of a frame according to the flow-input rate. When congestion occurs, frames are selectively discarded based on their color and the frame-dropping threshold. Through the notification process for the frame-dropping threshold, frames are discarded at upstream nodes in advance, avoiding the accumulation of a queuing delay. The performance of NRN+1CM was analyzed theoretically and evaluated with computer simulations. However, its ability to avoid bufferbloat has not yet been proven mathematically. This paper uses an M(n)/M/1/K queue model to demonstrate how bufferbloat is avoided with NRN+1CM's frame-dropping threshold-notification process. The M(n)/M/1/K queue is an M/M/1/K queuing system with balking. The state probabilities and average queue size of each ring node were calculated with the model, proving that the average queue size is suppressed in several frames, but not in the most congested queue. Computer simulation results confirm the validity of the queue model. Consequently, it was logically deducted from the proposed M(n)/M/1/K model that bufferbloat is successfully avoided with NRN+1CM independent of the network conditions including the number of nodes, buffer sizes, and the number and types of flows.

On the Similarities Between Urban Traffic Management and Communication Networks: Application of the Random Early Detection Algorithm for Self-Regulating Intersections

Communications are revolutionizing transport systems. Recently, novel technologies from the fields of telecommunications and data networks have been incorporated into both vehicle systems and road infrastructures, setting new frontiers in the design and development of Smart Cities and Smart Urban Mobility. Whereas each scenario, i.e., the vehicular one and the data communications one, has its own and well-defined particularities, there are also similarities that should be further explored. In this work, we address this confluence of research areas by investigating the applicability in vehicle traffic management of an active queue control algorithm employed in communication networks and called Random Early Detection (RED). RED provides congestion avoidance in data networks by controlling the average queue size of network buffers. It detects incipient congestion and, using probabilistic packet-marking or packet-dropping behavior, it is able to keep the data network at high throughput and low delay. In this work, we propose an adaption of RED to vehicle traffic management in signaled intersections, namely, RED for Vehicles (REDV). A complete characterization of REDV is provided, including the impact of several configuration parameters. We also conduct an extensive comparative performance evaluation via computer simulation. Results show that REDV is a light and simple algorithm able to significantly reduce the average trip time, waiting time, and CO2 emissions in a signaled intersection compared to other approaches from the related literature.

Gateway load balancing using multiple QoS parameters in a hybrid MANET

A Mobile ad hoc network (MANET) is a self configurable wireless network in which mobile nodes communicate with each other in a multihop fashion without any pre-installed infrastructure. A MANET can be considered to be a standalone network. To enhance the connectivity of a MANET it can be connected to the fixed network, thus forming a heterogeneous network. The integration of MANET and the Internet is called a hybrid MANET which is facilitated by special nodes called Internet gateway nodes. Load balancing among gateways is a challenging task when a MANET is connected to Internet. Gateway nodes with higher loads will lead to disconnected networks and depletes the node’s resources which include their batteries, memory and bandwidth quickly. Gateway selection based on the shortest path may increase traffic concentration on one particular gateway which leads to congestion and increases delay in the network. In this paper a QoS based load balancing mechanism has been proposed among multiple gateway nodes that provide communication between mobile nodes and fixed nodes in the Internet to select lightly loaded gateways so that more packets will be delivered to the fixed host in the Internet. The proposed QoS based scheme selects four QoS parameters that are (1) connecting degree, (2) interface queue length, (3) routing table entries and (4) hop count. A weight based method is used to select the gateway which combines all four QoS metrics. Simulation results demonstrate that when compared with individual parameter, the average ETE delay, queue size and traffic load of gateway generated by proposed algorithm is decreased by 17, 25 and 15 % respectively and when compared with existing schemes, the average ETE delay, queue size and traffic load of gateway is decreased by 25, 25 and 16 % respectively.

Adaptive mean queue size and its rate of change: queue management with random dropping

The random early detection active queue management (AQM) scheme uses the average queue size to calculate the dropping probability in terms of minimum and maximum thresholds. The effect of heavy load enhances the frequency of crossing the maximum threshold value resulting in frequent dropping of the packets. An adaptive queue management with random dropping algorithm is proposed which incorporates information not just about the average queue size but also the rate of change of the same. Introducing an adaptively changing threshold level that falls in between lower and upper thresholds, our algorithm demonstrates that these additional features significantly improve the system performance in terms of throughput, average queue size, utilization and queuing delay in relation to the existing AQM algorithms.

Energy-Aware Wireless Scheduling With Near-Optimal Backlog and Convergence Time Tradeoffs

This paper considers a wireless link with randomly arriving data that are queued and served over a time-varying channel. It is known that any algorithm that comes within ε of the minimum average power required for queue stability must incur average queue size at least Ω(log(1/ε)). However, the optimal convergence time is unknown. This paper develops a scheduling algorithm that, for any ε > 0, achieves the optimal O(log(1/ε)) average queue size tradeoff with a convergence time of O(log(1/ε)/ε). An example system is presented for which all algorithms require convergence time at least Ω(1/ε), and so the proposed algorithm is within a logarithmic factor of the optimal convergence time. The method uses the simple drift-plus-penalty technique with an improved convergence time analysis.

Applying Fuzzy Logic Principles to Improve the Performance of the Random Early Detection Algorithm

This paper proposes a Random early detection algorithm based on fuzzy logic Principles. The main target of using the fuzzy logic is to reduce the number of lost packets which are sent by a sender using RED algorithm in queue-buffer router of the network topology. The function of fuzzy logic is to dynamically tune the maximum drop probability (maxp) parameter of the RED algorithm. To realize this target, a twoinput-single-output fuzzy logic is implemented. The inputs of the fuzzy logic are average queue size, the difference in average queue size. To estimate the performance of the FLRED: simple network topology with FTP is suggested. In this research, the opnet modeler 14.5 has been used. The simulation results show that the FLRED algorithm is better than traditional RED algorithm as far as the number of lost packets is concerned.

High utilization dynamic bandwidth allocation algorithm based on sorting report messages with additive-polling thresholds in EPONs

High utilization is one of the design goals for MPCP-based dynamic bandwidth allocation (DBA) algorithms in Ethernet passive optical networks (EPON). However, utilization is sacrificed in order to meet the delay limits of the applications in most DBA design schemes. This paper proposes a dynamic bandwidth allocation algorithm based on sorting report messages with additive-polling thresholds (AP-Sort DBA). This has the characteristics of high utilization and low delay during medium network loading. AP-Sort DBA is an extension of the adaptive dynamic bandwidth allocation algorithm with sorting report messages (Sort-DBA) and promotes utilization by reduction of unused slot remainders (USR) and distribution of excess bandwidth. For performance evaluation in terms of average delay, average queue size, loss rate and utilization, five DBA schemes are simulated, namely AP-Sort DBA, interleaved polling with adaptive cycle time (IPACT), dynamic bandwidth allocation with a modified grant table generation algorithm and fair-excess allocation (DBA2-FE), Sort-DBA and double phase polling algorithm (DPA). Self-similar traffic and OC-3 packet size distribution are used for frame generation. The results show that AP-Sort DBA can achieve up to 99% utilization, which is the highest among the tested alternatives, with up to 60% improved delay compared to the IPACT. Technical analysis is presented for further proof of the effectiveness of the proposed methodology.

Enhanced uplink scheduling algorithm for efficient resource management in IEEE 802.16

Abstract Worldwide interoperability for Microwave Access (WiMAX) is based on the IEEE 802.16 standard. This technology provides broadband wireless last-mile access in a Metropolitan Area Network (MAN). The IEEE 802.16-2004 standard (i.e. Fixed WiMAX) provides specification for the Medium Access Control (MAC) and Physical (PHY) layers for WiMAX. A critical part of the MAC layer specification is scheduling, which resolves contention for bandwidth and determines the transmission order of users. The algorithms are researched under different mixes of traffic and for various characteristics of the IEEE 802.16 MAC layer. In this research the focus is on the WiMAX uplink traffic scheduling. A scheduling algorithm’s task in a multi-class network is also to categorize the users into one of the pre-defined classes. In this research, the algorithms are tailored at enhancing the collective performance of hybrid algorithms in the WiMAX domains which complement the assigned priorities. The spectrum of constraints which have been extracted from the hybrid Earliest Deadline First(EDF) + Weighted Fair Queuing (WFQ) + First in First out (FIFO) EDF+WFQ+FIFO algorithm include the static nature by which priorities are assigned and maintained during the entire duration of a transmission time and the core attributes of absolute dependence on deadline. The second constraint is embedded within the EDF scheduling algorithm and the perseverance of pursuing deadline associated weightages. In this research, the reengineering of the scheduling mechanics governing the EDF algorithms has been pursued. The dominance of the pre-stipulated deadline of the EDF algorithm is indeed acknowledged in the proposed and developed enhanced algorithm. The simulation results indicate that the legacy algorithms are not suitable for the multi-class traffic systems of WiMAX. This is because these algorithms do not explicitly incorporate the WiMAX QoS parameters into their mechanisms and are highly static. Extensive discrete – event simulation experiments have been done for the purpose of performance analysis. The performance metrics used are average throughput, average delay, missed deadline ratio and average queue size utilization ratio. The acquired results have proven that the proposed algorithms have successfully enhanced the static constrained algorithm.

Analysis of Batch Arrival Queue with Two Stages of Service and Phase Vactions

We study a batch arrival queueing system of phase vacation with two stages of service based on a Bernoulli schedule. A single server provides essential service to all arriving customers with service time following a general distribution. After two stages of service completion, the server leaves for phase one vacation of random length with probability $p$ or to continue staying in the system with probability $1-p$. As soon as the completion of phase one vacation, the server undergoes phase two vacation. On completion of two heterogeneous phases of vacation the server returns back to the system. The vacation times are assumed to be general. The server is interrupted and the service interruption follows an exponential distribution. The arrivals follow a Poisson distribution. Using supplementary variable technique, the Laplace transforms of time dependent probabilities of system state are derived. From this we deduce the steady state results. We also obtain the average queue size and average waiting time.

Proactive and Adaptive Data Migration in Hierarchical Storage Systems using Reinforcement Learning Agent

With the data generation rates growing exponentially, businesses are having a difficult time maintaining data center infrastructure. Hierarchical storage systems has evolved as a better alternate to managing data, as frequently accessed data is placed on higher tiers and the least frequently accessed data on lower tiers. But the data arrangement is not always static. Data Migration is an operation in which the selected data is physically moved, or migrated, to different storage components. The existing method of data migration in hierarchical storage systems has several shortcomings – reactive, heuristic policies and proprietary software. This paper attempts to overcome the above limitations by applying a reinforcement learning (RL) agent in the data migration of hierarchical storage systems. By the addition of RL agent the data migration which earlier was responsive, is made proactive and adaptive. Experimental results demonstrate the effectiveness of the proposed RL agent. The results indicate that: (i) the average queue size of storage devices is reduced to almost zero as the RL agent proactively migrates data and (ii) at 95 % confidence level the RL agent has no affect on the read and write operation of certain file sizes.

The Modified Gaussian Function based RED (MGF-RED) Algorithm for Congestion Avoidance in Mobile Ad Hoc Networks

In this paper the idea is to rationale a new active queue management technique (MGF-RED) based on Random Early Detection (RED), a well known active queue management technique for congestion avoidance. The conventional RED uses the concept of minimum threshold, maximum threshold and average queue length. If the average queue length increases by minimum threshold then the algorithm start dropping the packets with some probability using linear function of average queue size. If the average queue size is greater than the maximum threshold then packets are dropped with 1 probability. The proposed algorithm will use Modified Gaussian function for calculating the likelihood of packet drop as the average queue length increases the minimum threshold. Simulation has been performed using NS-2.34. The simulation results show that the proposed algorithm will present better network performance in throughput, packet delivery fraction and number of dropped packets. General Terms MANET, RED, Gaussian Function.

Reducing Packet loss in Active Queue Management

Random Early Detection (RED) is a network congestion control algorithm which detects incipient congestion according to current length of average queue. It detects congestion only after a packet has been dropped. In the recent, high speed networks, it is important to have some mechanisms that keep throughput high and average queue sizes low without packet loss and it would be undesirable to have large queue because it would significantly increase the average delay in the network.

Comparison of PI controllers designed for the delay model of TCP/AQM networks

One of the major problems of communication networks is congestion. In order to address this problem in TCP/IP networks, Active Queue Management (AQM) scheme is recommended. AQM aims to minimize the congestion by regulating the average queue size at the routers. To improve upon AQM, recently, several feedback control approaches were proposed. Among these approaches, PI controllers are gaining attention because of their simplicity and ease of implementation. In this paper, by utilizing the fluid-flow model of TCP networks, we study the PI controllers designed for TCP/AQM. We compare these controllers by first analyzing their robustness and fragility. Then, we implement these controllers in ns-2 platform and conduct simulation experiments to compare their performances in terms of queue length. Taken together, our results provide a guideline for choosing a PI controller for AQM given specific performance requirements.