Zhengxin Ma

Spatial multiuser access with MIMO smart antennas for OFDM systems

Indoor wireless communication systems have grown rapidly because of their clear advantages such as mobility, flexibility, and inexpensive network reconfiguration. In order to offer higher data rates approaching those provided by wired LANs, a combined OFDM/SDMA-based approach is an effective solution for increasing the system capacity and spectral efficiency. However, in multi-user environments, the system performance is limited by co-channel interference. Multi-input-multi-output (MIMO) smart antennas with prior knowledge of the channel at the transmitter is another promising technique for providing significant increase in system capacity and performance in wireless communication systems. We investigate the use of smart antennas at both the base and mobile stations, operating jointly, to maximize the SINR of each user before multiuser detection. By doing so, the performance of multi-user detection is significantly improved.

ML-oriented DA sampling clock synchronization for OFDM systems

The paper describes an error feedback loop for performing the data-aided (DA) sampling clock synchronization in orthogonal frequency division multiplexing (OFDM) systems. The algorithm is derived from the joint maximum likelihood (ML) estimation of the sample timing and the carrier phase in the AWGN channel. It is shown that the sample timing estimation variance of the proposed algorithm is asymptotically equal to the Cramer-Rao bound (CRB).

A stability-based link state updating mechanism for QoS routing

QoS routing, which satisfies diverse application requirements and optimizes network resource utilization, needs accurate link states to compute paths. Suitable link state update (LSU) algorithms which ensure timely propagation of link state information are thus critical. Since traffic fluctuation is one of the key reasons for link state uncertainty and existing approaches cannot effectively describe its statistical characteristics, we propose a novel stability-based (SB) LSU mechanism which consists of a second-moment-based triggering policy and a corresponding stability-based routing algorithm. They incorporate knowledge of link state stability in computing a stability measure for link metrics. With extensive simulations, we investigate the performance of the SB LSU mechanism and evaluate its effectiveness compared with existing approaches. Simulation results show that SB LSU can achieve good performance in terms of traffic rejection ratio, successful transmission ratio, efficient throughput and link state stability while maintaining a moderate volume of update traffic.

A mitigating stagnation-based ant colony optimization routing algorithm

To guarantee QoS requirement of traffics and utilize network resources efficiently have become the problem to be solved peremptorily in the next generation network (NGN). It has been researched that, as a simulation of natural ant, the artificial ant colony algorithm has shown its advantages over many previous ones. However, stagnation phenomenon has a serious impact on the performance and adaptivity of ant colony algorithm. Based on extensive survey, this paper proposes a mitigating stagnation-based ant colony optimization routing algorithm (MS-ACO), computer simulations and analysis indicate that, compared with other existing ant colony algorithms to mitigate stagnation, the proposed algorithm can obtain better QoS performances.

A multipath routing scheme combating with frequent topology changes in wireless ad hoc networks

The main challenge in mobile ad hoc networks is how to discover and maintain the routes to the destinations in the wireless environment with frequent dynamic topology changes, which results from the random movements of stations. According to the number of available paths to the destination, the prior proposed routings can be classified into unipath and multipath protocols. In this paper, to combat the frequent topology changes, we propose a neighbor-table-based multipath routing scheme which utilizes the advantages of multipath routings. In NTBMR, we introduce time-driven and data-driven mechanisms to set up the neighbor table. During route discovery, a joint route selection criterion which utilizes multiple attributes of the route is employed. Simulation results show that our multipath scheme can improve the network performance in terms of packet delivery ratio and end-to-end delay, which is important for real-time applications. However, the advantages are achieved at the cost of overhead about 7% higher than that of the unipath routing.

A Low Pass Filter Traffic Shaping Mechanism for the Arrival of Traffic

To decrease the traffic burstiness, general method of traditional shaping mechanisms is to smooth the traffic rate or the packet interarrival time of existing flows. However, traffic burstiness is also induced by the gusty arrivals of new flows. According to our knowledge, there is no research which attempts to decrease the burstiness caused by the arrivals of new flows. To cope with this, a shaping mechanism to the interarrival of flow named interarrival low pass filter (ILPF) fore-shaping mechanism is proposed in this paper. By the low pass filter, the ILPF fore-shaping mechanism smoothes the flow interarrival time so as to decrease the traffic burstiness. The ILPF fore- shaping mechanism is well suitable for the real-time applications which can endure some delay for accessing. According to theoretic analysis and computer simulation, it is proved that the ILPF fore-shaping mechanism significantly improves the utilization of the network as well as providing QoS guarantees in terms of probability.

Syndrome-based hybrid ARQ with Reed-Solomon codes and the practice in image transmission over Rayleigh fading channel

Two Reed-Solomon (RS) coded hybrid ARQ schemes based on syndrome calculation are proposed. For poor correction capacity RS code, the retransmission request is determined in the decoding procedure, which we call syndrome-ahead hybrid ARQ. As to powerful RS code, the request arises after decoding. The scheme thus gets the name of syndrome-behind hybrid ARQ. To guarantee the maximal throughput of the system, each retransmission is performed only once. We use analytical framework to evaluate the performance of the two schemes in Rayleigh fading channel. Simulation results are also provided to enhance the conclusion. The two hybrid ARQ scenarios are then applied to compressed image transmission, and the results further prove that the proposed ARQ schemes are powerful and effective to be used for a variety of fading channel applications.

Adaptive Call Admission Control for Real Time Video Communications Based on Delay Probability Distribution

In QoS networks, Connection Admission Control (CAC) is required to ensure that network resources are used efficiently. Very often CAC algorithms attempt to estimate available network resources (e.g., bandwidth, buffer space) to make admission decisions. In the case of real time video applications, transmission delay and packet loss rate constraints are crucial QoS parameters. However, the mapping between these parameters and available network resources is not so easy. In this paper, we present and analyze an adaptive admission control and traffic adjustment algorithm for real time video transmission, which is based on the measurement of end-to-end delay probability distribution and the user-defined QoS level profiles. As the network status varies, the algorithm attempts to improve or degrade the QoS levels of existing calls within the acceptable range. Simulation results demonstrate that the proposed algorithm performs well in terms of decreasing the call blocking rate while utilizing the network bandwidth effectively.

A Feedback Control Interval Shaping Algorithm under Delay Bound Restriction

Traffic shaping is an important technique for QoS guarantees in the network that provides integrated services. All shaping algorithms reduce traffic burstiness by introducing some shaping delay. However, how to achieve a trade-off between burstiness reduction and traffic delay is a challenging problem in current networks. Considering that the end-to- end delay bound is an important parameter in QoS guarantees, in this paper, we present a feedback control interval shaping algorithm (FCISA) which could shape the traffic effectively under a given delay bound restriction. Simulation results show that FCISA can achieve a satisfying performance.

An Improved Random Retreat CAC Algorithm for Interactive Real-time Applications

Traditional CAC methods all make decision as soon as possible on the arrival of a flow. However, the accessing delay is not required that strictly to some interactive real-time applications such as videoconference and IP telephone. In that case CAC should use this delay to find the appropriate moment for the flow to access network. Hence, in this paper we propose an improved CAC algorithm, namely Random Retreat CAC (RRCAC), based on an existing CAC framework. RRCAC refers the main idea of a carrier sense protocol-CSMA. If a flow cannot access the network due to the heavy load at the moment, RRCAC randomly assigns a retreat time to the flow until it is admitted or the delay exceeds a reasonable upper bound. The theoretic analysis and computer simulation show that the RRCAC algorithm significantly enhances the utilization of network and guarantees the QoS requirements in terms of probability.