Fei Ge

Throughput of FAST TCP in Asymmetric Networks

This letter proposes a model of FAST TCP that captures its dynamics in asymmetric networks. It is found that, as in symmetric networks FAST TCP can achieve stability but its throughput degrades as a function of a measure of the asymmetry that we call asymmetry factor. The theoretical results are validated by simulations.

HCEP: a hybrid cluster-based energy-efficient protocol for wireless sensor networks

This paper addresses the challenge of load balance and network lifetime extension in Wireless Sensor Networks (WSNs). We propose a novel distributed clustering approach for WSNs, which is termed as Hybrid Cluster-based Energy-efficient Protocol (HCEP) by using clustering as a key communication control technique. To measure the load distribution among nodes in a WSN, we quantify load balance by giving a load parameter for every node in a WSN. We then introduce a novel clustering function that takes into account a number of factors such as node degree, node residual energy and node initial energy. Based on the clustering function, we develop the framework of HCEP for implementation in WSNs. We design HCEP to incorporate two sorts of procedures in it, mainly the clustering procedures. Using the procedure, HCEP takes the advantage of load balance and uses localised coordination to enable energy efficiency and robustness for dynamic WSN networks, and thus it can prolong the network lifetime significantly. Simulation results verify that HCEP extends lifetime when communication radius changes within certain range and that the introduced balance parameter can be set to extend lifetime significantly.

Improving FAST TCP Performance in Asymmetric Networks

FAST TCP is a new promising TCP protocol developed for high-speed long-latency networks. This paper reveals that, within the FAST TCP protocol, the forward flow suffers performance degradation in asymmetric networks with bi-directional flows. We disclose the reason for the performance degradation, i.e., the queue delay measured by forward flow in one-way path is severely influenced by backward flow. A novel method is proposed to ameliorate this situation by using TCP time stamp option. Simulation results demonstrate significant performance improvement of forward flow.

On utility-optimised router-level bandwidth allocation

Kellys network utility maximisation (NUM) problems and solutions are aimed to maximise the aggregate utility subject to link capacity constraints. They are formulated and solved by using the individual flow rate vector. Because of the architecture of the current networks such as the Internet, the individual flow rates are generally not measurable directly at the routers for the network service provider. However, the aggregate flow rates are more convenient to obtain and to adjust. In this paper, we still study the NUM problems for communication networks but from a router-level bandwidth allocation standpoint. With the use of the generalised matrix inverse, we propose a general model of utility-optimised router-level bandwidth allocation and its solution, where the objective function and the constraints are formulated in terms of the aggregate flow rate vector rather than from the individual flow rate vector as in the usual NUM problem. We find that the new proposed models are equivalent to Kellys NUM model in the sense that they lead to the same optimum and their solutions satisfy the given routing scheme. We also discuss the special cases where the routing matrix is of full row rank and where there is one single-hop flow in every link in the network. We suggest a direct application to Internet Protocol-based virtual private network of the latter case. We present the mathematical models and solution procedures that lead to the utility-optimised aggregate flow rate vector and further illustrate them by numerical examples. We believe our approach is promising for deployment in communication networks. Copyright

A Partial Super Fast Recovery Algorithm for FAST TCP

This paper proposes an algorithm termed as partial super fast recovery (PSFR) for dealing with the packet loss due to random link error in FAST TCP. PSFR increases throughout and alleviates oscillations of router queue when random link error appears in equilibrium status. The implementation of PSFR is described and analyzed. NS2 simulation results confirm satisfactory performance of PSFR.

Traffic matrix estimation: Advanced-Tomogravity method based on a precise gravity model

Traffic matrix TM is extremely important in many networking operations. This paper proposes a novel approach of TM estimation in large-scale IP networks, termed as Advanced-Tomogravity method, which is based on a precise gravity model and the tomography method. First, the precise gravity model is proposed on the basis of the existing generalized gravity model by introducing a relativity factor vector parameter, which defines the relativity between the solution of the existing generalized gravity model and its real TM. The solution obtained from this precise gravity model is then refined by the basic model of the tomography method. By mathematical analysis, we give the explicit expression of the relativity factor vector parameter in the proposed precise gravity model by the Moore-Penrose inverse and the minimum-norm least-square solution. The vector parameter is subsequently determined with the aid of small amount of historical real data of TM. A general algorithm of the proposed approach is therefore designed. Finally, our approach is validated by simulation using the real data of the Abilene Network. The simulation results indicate that it reduces the relative errors to less than one-half, better tracks not only the dynamic fluctuations but also the overall mean behavior of traffic flow. Copyright

Performance effects of two-way FAST TCP

In this paper we study the performance of delay-based congestion control in the presence of queueing in the reverse path. Specifically, we consider FAST TCP in a single-bottleneck network by considering two scenarios, each corresponding to an equilibrium where a dierent direction of the bottleneck link is saturated. We argue that the local stability properties of FAST are unchanged by reverse trac, and present expressions for the throughput of each flow and link. Finally, we consider the eect of bidirectional trac on the dynamics of the number of flows in the network. We find conditions on the trac loads under which the bottleneck shifts between the forward

Network utility maximization in two-way flow scenario

A communication network usually has data packets and acknowledge (ACK) packets being transmitted in opposite directions. ACK packet flows may affect the performance of data packet flows, which is unfortunately not considered in the usual network utility maximization (NUM) model. This paper presents a NUM model in networks with two-way flows (NUMtw) by adding a routing matrix to cover ACK packet flows. The source rates are obtained by solving the dual model and the relation to the routing matrix of ACK packet flows is disclosed. Furthermore, the source rates in networks with one-way flows by the usual NUM model are compared to those in networks with two-way flows by the NUMtw model.

Latency of FAST TCP for HTTP Transactions

For HTTP over FAST TCP scenario, this letter presents a simple closed-form formula to estimate the latency, taking into account the network parameters such as packet size, link capacity, propagation delay and FAST TCP parameter settings. Analyzes supported by simulation results reveal the effects of network parameters and FAST TCP parameters on the FAST TCP/HTTP latency.

FAST TCP performance under perturbation imposed queueing delay in equilibrium

In this paper, we investigate the performance of FAST TCP under perturbation imposed queueing delay in equilibrium in dumbbell networks. We derive the transfer functions of the delay perturbation to the FAST TCP congestion window, queueing delay, and throughput. By using the transfer functions, the effects of impulse, step, and ramp delay perturbation on the congestion window, queueing delay, and throughput are theoretically analyzed. It is shown by the analysis that ramp delay perturbation makes FAST TCP unstable and step delay perturbation leads to constant errors, while impulse delay perturbation does not influence FAST TCP stability. The throughput of FAST TCP flows under different types of delay perturbations are simulated by ns-2. The simulations agree with our analytical results.