We propose a fast and reliable target Evolved Node B (TeNB) selection scheme for Long Term Evolution-Advanced (LTE-A) handover (HO). Selection of the TeNB in LTE-A HO procedure is based on signal strength requirements and thus may not be ideal. Also, during each measurement period the user equipment (UE) needs to perform scanning of multiple neighboring eNB (NeNB) frequencies resulting in delays too large for high speed services. In our scheme, the TeNB is chosen based on three different parameters: the orientation match, the current load (CL) of NeNBs and their received signal strengths (RSS). This selection based on three independent parameters makes it more reliable, enhancing HO performance. Moreover, as a UE only measures the NeNBs shortlisted based on these parameters, the measurement time is also reduced, which leads to a reduction in the overall HO time.

This paper proposes a mobile station (MS)-controlled fast MAC-layer handover (HO) scheme to reduce the HO latency in Mobile WiMAX. Based on the received signal strength (RSS) from any Base Station (BS), the MS can estimate its present distance from that BS. This capability allows the MS, through utilization of the broadcast messages by its Serving BS (SBS) and through few scannings of the Neighbouring BSs (NBS), to self-track its own direction of motion relative to the SBS and the NBSs and thus look ahead to perform an intelligent choice of the Target BS (TBS) for the impending HO. Simulation studies have shown a considerable reduction in the HO latency in the proposed method.

In multiowner wireless networks, access points (APs) are owned and operated by different administrations, leading to significant authentication delays during handoff between APs. We propose to exploit the trust between the owners of neighboring APs for reducing the authentication delay. In the proposed authentication scheme, neighboring APs that trust each other share the security key for the visiting node to avoid lengthy authentication routines each time the visiting node switches APs. The performance of the proposed trust-based authentication scheme is evaluated using a Markov model. Using numerical experiments, we first study a basic scenario where mobile nodes are not aware of the trust networks that exist in a given neighborhood. Subsequently, we consider an advanced scenario where a mobile node functionality is augmented to discover the trust network so as to minimize roaming beyond the trusted APs. We find that, even with the basic implementation, the average number of full authentications needed for a roaming mobile reduces linearly as the likelihood of two neighboring APs trusting each other increases. With the advanced implementation, our experiments show that quadratic reduction is achieved. The Markov model is validated using discrete event simulation.

CARM is a recent rate management technique designed to provide throughput differentiation among TCP aggregates, as may be required in QoS architectures such as DiffServ Assured Forwarding Per-Hop-Behavior. In this paper we derive the system model for a TCP aggregate that is DiffServ controlled, and use it for the analytical design of the PI controller employed in CARM. The model explicitly accounts for DiffServ parameters such as the finite token bucket depth and contract rate. The designed controller leads to fast convergence and has parameters that can be locally computed. The use of integral control action can lead to the phenomenon of integral windup in realistic network scenarios, so the controller design also includes an anti-integral-windup formulation. We present ns-2 simulation results that validate our design.

In this paper we analyze the medium access control (MAC) of fixed WiMAX with respect to best effort (BE) traffic. WiMAX employs a contention based method with exponential backoff to facilitate multiple access. A two-dimensional Markov chain can be used to accurately model the backoff procedure. From this we derived expressions for channel access probability, access delay, system throughput and resource usage. The accuracy of the model and its results are validated by simulations carried out on an IEEE 802.16d simulator. Constraints and limitations of the model are also discussed. Insight from this model can be used to optimize bandwidth usage and channel partitioning.

In this paper, we extend the analytical model for the TCP steady state throughput as a function of the network utilization factor, round trip time and packet drop rate for unlimited data transfer to capture the fast retransmit and recovery mechanisms. These mechanisms are frequently activated on connections over wireless links. We model TCP with a modified fast retransmit algorithm that allows packet recovery with smaller congestion window sizes than possible with TCP Reno or NewReno, thereby reducing the likelihood of timeouts. We also propose a further modification that dynamically adjusts its congestion window by considering the packet drop rate as the input parameter. Simulations with the OPNET tool demonstrate that the model is able to predict TCP performance accurately over a wide range of packet loss rates. Moreover, simulation results show the performance of the modified TCP scheme is markedly superior to that of standard TCP Reno over a UMTS network.

Scalable TCP, based on a multiplicative-increase multiplicative-decrease congestion avoidance algorithm, has been proposed recently to overcome the inability of Standard TCP to utilize the full bandwidth in high-speed networks. This paper employs a novel approach to derive a transfer-function model of Scalable TCP that is then employed in a control-theoretic design of RED-based Active Queue Management for such a network. Robust stability of the proposed scheme is established under prescribed conditions and the design is validated by ns2 simulations.

The per-hop-behaviors (PHB) proposed for use within the DiffServ architecture are designed to meet the demands of diverse Internet applications. An important aspect of one such PHB, assured forwarding (AF), is the distribution of excess bandwidth among aggregates with different network parameters in an over-provisioned network. The lack of control over excess bandwidth distribution can be an impediment for successful adoption of AF PHB, particularly for applications built over congestion reactive transport protocols like TCP. This paper presents a model that helps understand the dependence of excess bandwidth distribution on different network parameters, including the number of flows per aggregate that has so far been ignored in previous studies. Furthermore, considering the dynamic behavior of a DiffServ controlled TCP aggregate we show that the mechanisms of DiffServ complicate TCP dynamics and changes its convergence properties. This knowledge can guide the design of networks and their various components such as meters and markers to achieve more control over the excess bandwidth distribution and make AF PHB more attractive. The model is validated through ns-2 simulation studies

Normal TCP performs well in wired networks. However, when employed in ad hoc wireless networks, its loss-based congestion window progression causes high network buffer utilization due to large bursts of data, which degrades the network bandwidth utilization. This paper argues that the reactive congestion window progression (based on duplicate ACK and timeouts) should be not used in ad hoc networks and proposes a receiver-oriented rate controller (rater), with a congestion window delimiter for the IEEE 802.11 MAC protocol, instead. In addition, the transient nature of medium availability due to medium contention is addressed by a freezing timer (freezer) at the receiver, which freezes the sender whenever heavy contention is perceived. Finally, ad hoc sender enhancements are proposed for optimizing the performance of the receiver-end, as an optional deployment. The new ad hoc TCP demonstrates outstanding results in terms of goodput, as well as throughput.

Useless packet transmission (UPT) arises when a multimedia stream becomes unintelligible because too many packets, in aggregate, are dropped by fair queuing algorithms at routers, even with low drop rates at individual routers. We propose two UPT avoidance (UPTA) schemes for networks with multiple congested links: partial UPTA (P-UPTA), which is core-stateless, and centralized UPTA (C-UPTA), which employs a bottleneck fairshare discovery (BFD) protocol to determine a flow's global fairshare. Our simulation study shows that P-UPTA eventually detects UPT, but bandwidth may be wasted on upstream links before UPT is detected, whereas C-UPTA avoids UPT in all situations as it always drops useless packets at the network edge. We quantitatively analyse the performance of C-UPTA in conjunction with WFQ, in terms of TCP throughput, file download time, MPEG-2 video intelligibility, and fairness. The results reveal that, for the scenarios simulated, the TCP throughput is improved by up to 50% without any significant impact on the intelligibility of the MPEG-2 video and the fairness of the scheduling algorithm.

This paper presents an analytical model that enables network dimensioning and configuration investigations for a general packet radio service (GPRS) network. The model accounts for transmission control protocol (TCP) using a detailed mathematical representation and introduces a 'TCP factor' metric and procedure to model TCP behavior. The analytical model results, for a web browsing application, are validated by means of discrete event simulation. The ability of the analytical model to perform accurate network dimensioning and admission control is also demonstrated.

The paper is concerned with optimizing end-to-end window controls that achieve proportional fairness in the long run, and then investigating their transient or short-term fairness. An abstracted stochastic model of a bottlenecked connection is employed and the window control is designed to minimize the buffer queue variance while keeping the mean queue level at a target value, taking the round-trip delay into account. A generalized window control aimed at reducing window size fluctuations is also derived. The effects of both the target queue length and a weighting parameter on the short-term fairness, as measured by short-term fairness indices calculated over short time intervals, are investigated using ns-2 simulations.

The route computing overhead caused by on-demand calculation of QoS routes, especially in large networks with heavy traffic, is a concern and can cause scalability problems. This paper introduces a distributed cache architecture to reduce the route computing load, caused by execution of QoS routing algorithms. The distributed nature of the proposed cache architecture facilitates its deployment in large networks. To maximize the performance of the distributed cache architecture, cache snooping has been proposed to alleviate the effects of changes in the network states. In performance evaluation of the proposed distributed cache architecture, we use a broad range of realistic network topologies, network traffic conditions, routing protocols, and aggregation techniques to evaluate different aspects of the proposed cache architecture under different conditions. The results confirm that the distributed cache architecture can significantly reduce the route computing overhead. In addition, our results suggest that the cache snooping can significantly increase the overall routing performance, especially in the presence of highly inaccurate network state information.

One problem with a RED router with fixed parameter settings is that the buffer queue length varies markedly with the traffic level. In this paper, an algorithm is derived analytically for automatically tuning the key RED parameter max/sub p/ to keep the average queue length at the target value despite a varying traffic load. A control-theoretic stability analysis is performed to obtain guidelines for setting the remaining RED parameters and the adaptation frequency. Simulation results show that the new adaptive algorithm is faster acting and more accurate than previously proposed heuristic methods.

The design of rate allocation and queue length control in computer networks is treated as a stochastic optimal control problem. The performance index is chosen to achieve the twin objectives of minimising queue length fluctuations and fully utilising the available bandwidth. Simple, practically realisable optimal control schemes are obtained for both LANs and WANs. An adaptive scheme is proposed where the auto-regressive parameters of the traffic, needed for gain calculations, are estimated by an LMS algorithm. Discrete event simulations are carried out to verify the fluid-flow models used in developing the controllers, to compare their performance against PI controllers proposed previously, and to study the effect of self-similar traffic. Two key results are obtained. First, queue-length fluctuations, and hence potentially packet losses, are much smaller for the optimal feedforward controller than for the PI controller. Second, in contrast to uncontrolled queues, the queue length variance decreases with increasing Hurst parameter for self-similar traffic.

Predicting traffic requirements is a fundamental objective of network management algorithms, whether explicitly stated or not. Considering the abundant evidence that traffic is long-range dependent (LRD) and the fact that the history of long-range dependent processes has significant impact on the present value of the process, it is natural to assume that predicting LRD traffic would be rather rewarding. Although there is indeed significant performance gain in using the correlation structure as the Hurst parameter increases, we show in this paper that this is primarily due to the utilization of the specific short-term correlations that occur within the structure of the traffic, rather than storing a long history of the traffic and utilizing the long-term correlations that exist between the future traffic level and this stored history. This conclusion is demonstrated using two specific management functions: rate control for elastic connections, and queue management using probabilistic packet discarding.

This paper proposes the concept of cumulative inter-ADU (application data unit) jitter for use in studying the performance of a multimedia system. The concept models the ADU arrivals from a network by keeping the running total of the differences between inter-ADU spacings at the arrival instants and the generation instants. For a given playback delay, the concept can be used to estimate the number of late ADUs. This estimation is done at the entry to the playback buffer rather than at the exit as with existing media synchronization. Extensive simulation results have shown that the ADUs indicated as late by the cumulative inter-ADU jitter are indeed discarded by the playback algorithm. We illustrate its applications by examples. In particular, we highlight the significance of the concept in improving media synchronization algorithms and network-level controls from an application-level perspective.

Mobile IP is an IETF RFC that supports transparent host migration on the Internet. In the transition to IP-based cellular networks, it is widely agreed that mobile IP provides an elegant solution for inter-domain, or macro-mobility management, but lacks critical aspects of mobility management needed within cellular networks. The shortcomings have led to the emergence of micro-mobility management protocols. This paper presents an IP-based mobility management protocol, /spl mu/-mobile IP, based on two enhancements, proactive handoff and paging support, to an Internet draft proposal, Regional Registrations. A signaling cost analysis is used to justify the need for such enhancements. We show through simulations the protocol's suitability for supporting real-time communications.

Long-range dependence is regarded as a fundamental property of network traffic. Using an original approach, we incorporate this property in a traffic control mechanism for elastic connections that can adapt to the instantaneous network load in a differentiated services-type framework. In this scenario, the network makes predictions of bandwidth requirements of the high-priority traffic and returns feedback information to the elastic source. We include a prediction compensation algorithm that compensates for the larger prediction errors for connections with longer roundtrip delay, and analyze the performance of this algorithm. The specific topology involved in traffic control for differentiated services is thus harnessed, together with the long-range dependence, to improve network performance, thereby counteracting the undesirable characteristics of self-similarity. Furthermore, an adaptive version of the rate-based control algorithm is studied, based on the use of real-time estimates of traffic parameters, including the mean, variance, and Hurst parameter.

This paper presents an ABR control scheme (ABR/sup 2/-DP) which maximises the utilisation of information available about the network traffic. Firstly, it takes into account the long-range dependence of the self-similar background network traffic, and utilises it for more accurate multi-step predictions of the explicit rate (ER). Secondly, prediction errors for connections with greater network delay are compensated for in the ER calculations for connections with smaller delay. Hence, the available bandwidth is distributed both in temporal and spatial dimensions. The scheme allows a reduction in buffer requirements and queueing delay. Performance results are presented and compared with a previously proposed scheme which uses digital filtering.

Considering the limited bandwidth of the wireless link, it is important that the error control mechanism in wireless networks be spectrally efficient. Towards this end, we develop a hybrid error control architecture which takes into account the use of hierarchical video coding. Additionally, the architecture also includes a module which estimates the packet error rate and round trip time observed by the receiver and adjusts the level of redundancy used based on the estimate. By choosing different options in the architecture, we get different error control schemes. In this paper, we investigate the performance of five different error control schemes, namely ARQ, pure FEC, hybrid FEC/ARQ, hybrid FEC/ARQ with priority-dependent redundancy, and adaptive hybrid FEC/ARQ with priority-dependent redundancy. We evaluate the performance of these schemes using MPEG-2 video traces. The results show that pure FEC offers the worst overall performance. The ARQ scheme offers the best performance under low bit error rate and short round trip time, while the priority-aware hybrid FEC/ARQ with or without FEC adaptation offers the best performance under other conditions.

Network performance, for the engineer designing traffic management protocols, is the key issue as it ensures that the network is functioning within its safe operating limits and that users of various applications are satisfied with the service received. Thus, in the face of uncertainty regarding the true nature of the system and traffic traversing the system, robustness entails maximising the performance of the network across the entire range of possibilities. We investigate standard robust prediction techniques for the design of ABR rate control mechanisms, but, while these optimize for the case with the greatest uncertainty, the system performance can suffer significantly for other cases. By introducing a spectral density comparison term to the entropy, we arrive at a performance measure which achieves robustness in the sense of maintaining high performance under uncertainty. Performance also includes computational efficiency, and various long-range dependent models are compared with regards to the robustness of the system, allowing the model with the least computational expense to be selected.

In currently proposed asynchronous transfer mode (ATM) congestion control schemes, the policing function is carried out at the network edge, which opens the possibility of cells being discarded or marked without reference to the actual message. This can lead to degradation of the service quality in the voice service or multiplication of the cell loss probability in the data service. The authors propose an approach to the policing function, called user-network policer, to resolve these problems. The scheme consists of a service-dependent user policer and a service-independent network policer. Users are responsible for policing and marking their traffic appropriately before sending them into the network. The network is only responsible for verifying the correctness of user policing and for transporting cells transparently across the network.>