On fluid‐flow modeling of priority based request scheduling for finite buffer SIP server

With the widespread deployments of voice‐over‐internet protocol services, the existing session initiation protocol (SIP) design cannot scale up for large network sizes. Events triggering a demand burst or a server slowdown can cause SIP server overload, overload propagation, and crash, thus bringing down the whole SIP network. Since the SIP retransmission mechanism exacerbates the overload condition, existing models created for a stable SIP system cannot be effectively used to analyze an overloaded server. In this paper, we propose a fluid‐flow model to characterize the behavior of the finite buffer SIP server equipped with priority‐based request scheduling mechanism (PRSM). The model for the PRSM uses primary and secondary queues for the original request messages and the retransmitted requests, respectively. The performance metrics, namely, the failed call attempts and the response delay from sending INVITE request until receiving a 100‐Trying response, are derived using the arrival time‐slot tracking and the removal processes of the proposed fluid‐flow model. We conducted test cases under the heavy traffic conditions, where the overload is caused by bulk and bursty arrivals or server slowdown. The numerical results closely match with the simulation results for all experiments, indicating that the proposed model can accurately capture the dynamic behavior of an SIP server with the PRSM. The experiments demonstrate that the number of failed call attempts is close to 0 and the mean response delay is kept constant around 175 ms for the PRSM when the buffer size is higher than 1K while both metrics are significantly higher for the conventional SIP.     

Local and distributed SIP overload control solution improving sustainability of SIP networks

Carriers have adopted session initiation protocol (SIP) in their next generation networks. Providing carrier‐grade service requires high availability in times of component failures, avalanche restart, flash crowds and denial of service attacks, which cause overload on SIP servers. Throughput of SIP servers is largely degraded during overload. We propose an SIP overload control (SIP‐OC) solution for local and remote situations, working in hop‐by‐hop and end‐to‐end modes. Our local SIP‐OC method uses a cross‐layer approach with negligible performance impact, while the implicit nature of our remote SIP‐OC allows detection of sophisticated overload conditions such as those caused by non‐SIP entities. Our remote SIP‐OC uses transaction response time as the basis for implicit overload detection. Coupling our local and remote SIP‐OC schemes, we show that the range of ‘sustainable’ overload that can be imposed on the system improves significantly. Moreover, incorporating a 2‐means filtering mechanism into our SIP‐OC scheme makes it perform well under packet‐loss. We also show that our proposed solution is robust to network latency and SIP server capacity fluctuations. All of our results are obtained from experiments over SIP testbeds including an experimental IP multimedia subsystem. Copyright © 2016 John Wiley & Sons, Ltd.     

Design,implementation and performance evaluation of a proactive overload control mechanism for networks of SIP servers

The extent and diversity of systems, provided by IP networks, have made various technologies approach integrating different types of access networks and convert to the next generation network (NGN). The session initiation protocol (SIP) with respect to facilities such as being in text form, end-to-end connection, independence from the type of transmitted data, and support various forms of transmission, is an appropriate choice for signalling protocol in order to make connection between two IP network users. These advantages have made SIP be considered as a signalling protocol in IP multimedia subsystem (IMS), a proposed signalling platform for NGNs. Despite having all these advantages, SIP protocol lacks appropriate mechanism for addressing overload causing serious problems for SIP servers. SIP overload occurs when a SIP server does not have enough resources to process messages. The fact is that the performance of SIP servers is largely degraded during overload periods because of the retransmission mechanism of SIP. In this paper, we propose an advanced mechanism, which is an improved method of the windows based overload control in RFC 6357. In the windows based overload control method, the window is used to limit the amount of message generated by SIP proxy server. A distributed adaptive window-based overload control algorithm, which does not use explicit feedback from the downstream server, is proposed. The number of confirmation messages is used as a measure of the downstream server load. Thus, the proposed algorithm does not impose any additional complexity or processing on the downstream server, which is overloaded, making it a robust approach. Our proposed algorithm is developed and implemented based on an open source proxy. The results of evaluation show that proposed method could maintain the throughput close to the theoretical throughput, practically and fairly. As we know, this is the only SIP overload control mechanism, which is implemented on a real platform without using explicit feedback.     

Applying control theoretic approach to mitigate SIP overload

The Session Initiation Protocol (SIP) retransmission mechanism is designed to maintain reliable transmission over lossy or faulty network conditions. However, the retransmission can amplify the traffic overload faced by the SIP servers. In this paper, by modeling the interaction between an overloaded downstream server and its upstream server as a feedback control system, we propose two Proportional-Integral (PI) control algorithms to mitigate the overload by regulating the retransmission rate in the upstream server. We provide the design guidelines for both overload control algorithms to ensure the system stability. Our OPNET^® simulation results demonstrate that: (1) without the control algorithm applied, the overload at a downstream server may propagate to its upstream servers and cause widespread network failure; (2) in case of short-term overload, both proposed feedback control solutions can mitigate the overload effectively without rejecting calls or reducing resource utilization, thus avoiding the disadvantages of existing overload control solutions for SIP networks.     

Expedited forwarding end‐to‐end delay and jitter in DiffServ

The scheduling disciplines and active buffer management represent the main components employed in the differentiated services (DiffServ) data plane, which provide qualitative per‐hop behaviors corresponding to the QoS required by supported traffic classes. In the first part of this paper, we compute the per‐hop delay bound that should be guaranteed by the different multiservice scheduling disciplines, so that the end‐to‐end (e2e) delay required by expedited forwarding (EF) traffic can be guaranteed. Consequently, we derive the e2e delay bound of EF traffic served by priority queuing–weighted fair queuing (PQ–WFQ) at every hop along its routing path. Although real‐time flows are principally offered EF service class, some simulations on DiffServ‐enabled network show that these flows suffer from delay jitter and they are negatively impacted by lower priority traffic. In the second part of this paper, we clarify the passive impact of delay jitter on EF traffic, where EF flows are represented by renewal periodic ON–OFF flows, and the background (BG) flows are characterized by the Poisson process. We analyze through different scenarios the jitter effects of these BG flows on EF flow patterns when they are served by a single class scheduling discipline, such as first‐input first‐output, and a multiclass or multiservice scheduling discipline, such as static priority service discipline. As a result, we have found out that the EF per‐hop behaviors (PHBs) configuration according to RFCs 2598 and 3246 (IETF RFC 2598, June 1999; RFC 3246, IETF, March 2002) cannot stand alone in guaranteeing the delay jitter required by EF flows. Therefore, playout buffers must be added to DiffServ‐enabled networks for handling delay jitter problem that suffers from EF flows. Copyright © 2008 John Wiley & Sons, Ltd.     

A load scheduler for SIP proxy servers: design,implementation and evaluation of a history weighted window approach

The widespread use of Session Initiation Protocol as a signalling protocol has created various challenges. An important one is that its throughput can be severely degraded when an overload happens in the proxy server because of several retransmissions from the user agent. One common approach to overcome this problem is ‘load balancing’. A balancer needs to know the status of proxy servers, which are continuously gathered implicitly or explicitly. Implicit methods have averagely less overhead than explicit ones. This paper attempts to prevent throughput reduction by balancing the loads among available proxy servers properly using an implicit mechanism called History Weighted Average Response time . The proposed algorithm is robust because it incurs no extra processing to proxy servers. The novelty of the mechanism is making use of ‘response time history’ to estimate the load being currently processed on servers. By implementing in a real testbed, throughput and scalability are improved compared with an important state‐of‐the‐art similar algorithm. This improvement stems from no need for modification in SIP protocol, easy implementation and application, simple computations for making decision and no need for extra feedback between servers and load balancer. Copyright © 2015 John Wiley & Sons, Ltd.     

A fragment‐based retransmission scheme with quality‐of‐service considerations for wireless networks

In order to satisfy quality‐of‐service requirements for real‐time multimedia applications over wireless networks, IEEE 802.11e has been proposed to enhance wireless‐access functionalities. In IEEE 802.11e, collisions occur frequently as the system load becomes heavy, and then, the latency for successfully transmitting data is lengthened seriously because of contention, handshaking, and backoff overheads for collision avoidance. In this paper, we propose a fragment‐based retransmission (FBR) scheme with quality‐of‐service considerations for IEEE 802.11e‐based wireless local area networks. Our FBR can be used in all proposed fragmentation‐based schemes and greatly decrease redundant transmission overheads. By utilizing FBR, the retransmission delay will be significantly improved to conform strict time requirements for real‐time multimedia applications. We develop an analytical model and a simulation model to investigate the performance of FBR. The capability of the proposed scheme is evaluated by a series of simulations, for which we have encouraging results. Copyright © 2011 John Wiley & Sons, Ltd.     

HSG‐ad hoc network: A novel hierarchical star graph ad hoc network with self‐organization and routing discovery free

In ad hoc wireless networks, most data are delivered by multi‐hop routing (hop by hop). This approach may cause long delay and a high routing overhead regardless of which routing protocol is used. To mitigate this inherent characteristic, this work presents a novel ad hoc network structure that adopts dual‐card‐mode, self‐organization with specific IP naming and channel assignment to form a hierarchical star graph ad hoc network (HSG‐ad hoc). This network not only expedites data transmission but also eliminates the route discovery procedure during data transmission. Therefore, the overall network reliability and stability are significantly improved. Simulation results show that the proposed approach achieves substantial improvements over DSDV, AODV, and DSR in terms of average end‐to‐end delay, throughput, and packet delivery ratio. Copyright © 2009 John Wiley & Sons, Ltd.     

Bio-inspired auto-adaptive SIP overload controller

In this article, a Session Initiation Protocol (SIP) overload control solution is proposed. It considers all the types of SIP requests. This is really what a SIP load is composed of, in an industrial environment. So far, the specialized literature considered INVITE messages only. So, we think that SIP servers are required to be dynamically adaptive to the diversity of the incoming load content. In the latter, the rate of a given SIP message type may be more or less than the other message types, depending on the services provided by the SIP server. Sometimes, it also depends on the time of the day. The auto-adaptation ability of the proposed overload control mechanism is designed after the immune system metaphor. The solution is validated through load tests and compared with a well known SIP overload control algorithm. Test load arrival patterns have been chosen to simulate three different service packages known in the SIP industry world as: Hosted Private Branch Exchange, Prepaid Calling Card Service, and Call-Shop Service.     

Using GI‐G‐1 queuing model for rtPS performance evaluation in 802.16 networks

Among the scheduling services, rtPS (real‐time polling service) is designated for real‐time applications. Among three packet delay intervals, performance effect on polling interval has been widely studied, but less on the intervals of scheduling and delivery. To evaluate the performance of delay‐sensitive rtPS applications, instead of using continuous queueing model, a discrete‐time GI‐G‐1 model, which considers intervals of polling, scheduling, and delivery, is proposed. By taking VoIP as a typical rtPS application, the transmission latency under different QoS settings, polling probability, and traffic load are presented. The latency is also compared among various codec schemes. The results indicate that when the codec rate is either fulfilled or dissatisfied by the promised bandwidth of service levels, the performance is highly dependent upon the polling probability, no matter what the traffic condition is. However, if the codec rate is in between the promised bandwidth of various service levels, the polling probability is a dominant factor in light traffic environment, while the settings on QoS parameters will strongly determine the performance in heavy traffic situation. In addition to the verification using simulation, the bandwidth utilization derived from the GI‐G‐1 model can be applied to improve the serving capacity of base stations. Copyright © 2011 John Wiley & Sons, Ltd.     

Performance analysis of the internet‐protocol multimedia‐subsystem's control layer using a detailed queueing Petri‐net model

In the current trend in telecommunications industry towards all‐internet‐protocol (IP) infrastructures, IP multimedia subsystem (IMS) plays a critical role by providing a coherent data and control‐plane solution for large‐scale live multimedia applications in a flexible and cost‐effective manner. On the other hand, such a large‐scale service platform would inevitably fail without effective support for the quality‐of‐service (QoS) requirements perceived by its users. Among the most important factors that influence user QoS are system performance and scalability. In this paper, a performance model for IMS systems is developed using the queueing Petri nets (QPNs) as the modeling formalism. The model's parameters are tuned based on the measurements carried out using a well‐known IMS implementation. The model is validated against the real system. During the model calibration, the Java garbage‐collector process used in the home subscriber‐server (HSS) implementation was found to be a main factor in the discrepancy between the model and the reality. In addition, the effects of other factors such as the network stack in the operating system are investigated. The validated model is employed to give insights into the scalability of every single instance of IMS implementation. The model is extended to study load balancing among multiple instances of HSS to remove the main bottleneck in the system. It provides a valuable platform for resource management of various components of the IMS ecosystem to support the intended level of QoS for the users.     

An efficient MAC protocol with cooperative retransmission in mobile ad hoc networks

In emerging wireless networks, cooperative retransmission is employed to replace packet retransmission between a pair of sender and receiver with poor channel condition. A cooperative MAC protocol which utilizes such benefit is proposed in this paper to improve the network performance in mobile ad hoc networks. In the proposed protocol, relay nodes between sender and receiver are used if the sender cannot communicate with the receiver reliably. Furthermore, the receiver may also stop forwarding the received data frame if the frame is received by the next‐hop receiver on the route to the final destination node. Simulation results show that the proposed protocol outperforms previous works in terms of increased transmission reliability and reduced delay time. Copyright © 2010 John Wiley & Sons, Ltd.     

Analyses of soft‐state signaling protocols in GMPLS‐Based WDM networks

Wavelength‐routed Generalized Multi‐Protocol Label Switching (GMPLS) networks use Resource reSerVation Protocol—Traffic Engineering (RSVP‐TE) as signaling protocol to set up and tear down lightpaths. RSVP‐TE uses a soft‐state control mechanism to manage lightpaths. In the soft‐state control mechanism, each node sets a timer for each control state and resets the timer with refresh messages to maintain the state. When the timer expires due to losses of refresh messages, the control state is initialized and a reserved resource managed with the state is released. It has been considered that resource utilization of soft‐state protocols is inferior to that of hard‐state protocols, since soft‐state protocols may reserve resources until control states are deleted due to timeout. Therefore, some extensions to promote the performance of soft‐state protocols, such as message retransmission, have been considered. In this paper, we analyze the behavior of GMPLS RSVP‐TE and its variants with a Markov model and analyze the performance of RSVP‐TE. From the results, we demonstrate that resource utilization of RSVP‐TE can be equivalent to that of a hard‐state protocol when the loss probability of signaling messages is low. We also investigate the effectiveness of message retransmission and show that using message retransmission leads to poor resource utilization in some cases. Copyright © 2012 John Wiley & Sons, Ltd.     

Optimal control of a removable server in an M/Ek/1 queueing system with finite capacity

This paper deals with the economic behavior of a removable server in the N policy M/E_k/1 queueing system with finite capacity. Expressions for the probability mass functions of the number of customers in the system are derived and taken in closed-form. As special cases, the probability mass functions of the number of customers for the N policy M/M/1 queueing system, the ordinary M/K_k/1 queueing system, and the ordinary M/M/1 queueing system are obtained. The cost structure includes a holding cost per unit time spent in the system for each customer, costs per unit time for keeping the server on or off, a server start-up cost, a server shut-down cost, and fixed cost for every lost customer. Following the construction of the total expected cost per unit time, we determine the optimal operating policy at minimum cost.     

Rack‐Level DC Power Solution for Volume Servers

Rack‐level DC power supply is the optimal technology for providing DC power to a volume server without any power infrastructure changes in an existing AC data center. In this paper, we propose a smartly controllable and monitorable DC rack power system. The proposed system improves power efficiency by changing the power distribution architecture of a conventional method in the rack. We developed an optimal power control method in multipower modules to provide high efficiency at low loads. In addition, the proposed system provides real‐time web monitoring of the rack power and environment around a rack. In our experiments, the proposed system improved the power efficiency by over 10% compared to an AC power system providing N+1 redundant power and power monitoring.     

Finite Buffer Behavior with Poisson Arrivals and Random Server Interruptions

A queueing model with finite buffer size, Poisson arrival process, synchronous transmission and server interruptions is studied through a Bernoulli sequence of independent random variables. An integrated digital voice-data system with Synchronous Time-Division Multiplexing (STDM) for voice sources and Poisson arrival process for data messages is considered as an application for this model. The relationships among overflow probabilities, buffer size and expected queueing delay due to buffering for various traffic intensities are obtained. The results of this study are portrayed on graphs and may be used as guide lines for the buffer design in digital voice-data systems. Although this study arose in the design of a buffer for digital voice-data systems, the queueing model developed is quite general and may be useful for other industrial applications.     

Proxy‐based failure detection in multimedia streaming environments

This paper presents an effective failure‐detection mechanism called the ‘dynamic threshold,’ which has specifically developed for use with multimedia streaming applications. Compared to the conventional failure‐detection methods, which normally use a fixed threshold to detect a server failure, the proposed mechanism detects the failure in an effective and timely manner by using a dynamically adjusted timeout value upon which queries are sent to suspected streaming servers running in a dynamic network environment. The failover mechanism based on the dynamic threshold was implemented in a real streaming environment. The empirical results show that the dynamic threshold mechanism performs better than the conventional method in terms of timeliness, with comparable accuracy. Copyright © 2006 John Wiley & Sons, Ltd.     

A hop‐by‐hop delay‐constrained routing algorithm with explicit loop avoidance and backup routing information

QoS Routing is crucial for QoS provisioning in high‐speed networks. In general, QoS routing can be classified into two paradigms: source routing and hop‐by‐hop routing. In source routing, the entire path to the destination node of a communication request is locally computed at the source node based on the global state that it maintains, which does not scale well to large networks. In hop‐by‐hop routing, a path‐selecting process is shared among intermediate nodes between the source node and the destination node, which can largely improve the protocol scalability. In this paper, we present the design of hop‐by‐hop routing with backup route information such that each intermediate node can recursively update the best known feasible path, if possible, by collectively utilizing the routing information gathered thus far and the information that it locally stores. Such a route is kept as a backup route and its path cost is used as a reference to guide the subsequent routing process to search for a lower‐cost constrained path and avoid performance degradation. In this way, the information gathered is maximally utilized for improved performance. We prove the correctness of our presented algorithm and deduce its worst message complexity to be O(∣V∣²), where ∣V∣ is the number of network nodes. Simulation results indicate that, however, the designed algorithm requires much fewer messages on average. Therefore it scales well with respect to the network size. Moreover, simulation results demonstrate that the cost performance of our algorithm is near‐optimal. Copyright © 2004 John Wiley & Sons, Ltd.     

A novel efficient power‐saving MAC protocol for multi‐hop MANETs

Following recent advances in the performance of ad hoc networks, the limited life of batteries in mobile devices poses a bottleneck in their development. Consequently, how to minimize power consumption in the Medium Access Control (MAC) layer of ad hoc networks is an essential issue. The power‐saving mode (PSM) of IEEE 802.11 involves the Timing Synchronization Function to reduce power consumption for single‐hop mobile ad hoc networks (MANETs). However, the IEEE 802.11 PSM is known to result in unnecessary energy consumption as well as the problems of overheating and back‐off time delay. Hence, this study presents an efficient power‐saving MAC protocol, called p‐MANET, based on a Multi‐hop Time Synchronization Protocol, which involves a hibernation mechanism, a beacon inhibition mechanism, and a low‐latency next‐hop selection mechanism for general‐purpose multi‐hop MANETs. The main purposes of the p‐MANET protocol are to reduce significantly the power consumption and the transmission latency. In the hibernation mechanism, each p‐MANET node needs only to wake up during one out of every N beacon interval, where N is the number of beacon intervals in a cycle. Thus, efficient power consumption is achieved. Furthermore, a beacon inhibition mechanism is proposed to prevent the beacon storm problem that is caused by synchronization and neighbor discovery messages. Finally, the low‐latency next‐hop selection mechanism is designed to yield low transmission latency. Each p‐MANET node is aware of the active beacon intervals of its neighbors by using a hash function, such that it can easily forward packets to a neighbor in active mode or with the least remaining time to wake up. As a consequence, upper‐layer routing protocols can cooperate with p‐MANET to select the next‐hop neighbor with the best forwarding delay. To verify the proposed design and demonstrate the favorable performance of the proposed p‐MANET, we present the theoretical analysis related to p‐MANET and also perform experimental simulations. The numerical results show that p‐MANET reduces power consumption and routing latency and performs well in extending lifetime with a small neighbor discovery time. Copyright © 2011 John Wiley & Sons, Ltd.     

SIP视频会议中服务器的研究与改进

视频会议系统中会议服务器决定着会议的稳定性、规模和质量.通过研究SIP协议、SIP协议实体和SIP视频会议中服务器的结构等,对现有结构和服务器进行了改进和设计.采用负载均衡方法,提高了系统动态自适应能力,扩大了视频会议的规模.利用SIP消息分发以及优先处理机制,进一步提高了视频会议服务器的消息处理效率,以解决消息并发和消息拥塞的问题,提高系统的可靠性.