- EI核心期刊
Mobile ubiquitous service environment was put forward here as the vision of future mobile wireless world. The novel network evolution philosophy for such a vision was promoted. Several challenges were also discussed, such as the mutual affection of hetero organization and self-organization, end to end reconfiguration and the cross platform synthetic service construction, etc.
The delay performance of the indirect handoff scheme (IHS) experienced by constant bit rate (CBR) traffic was analyzed first, and the result shows this scheme in cellular IP (CIP) protocol cannot guarantee zero packet loss during the handoff process. Slow route updating and long access time to the new base station (BS) are the two main reasons leading to the potential packet loss. In order to mitigate the influence of these two problems, an enhanced indirect handoff scheme (EIHS) was proposed. Simplified routing updating (SRU) and handoff buffering (HBUF) were adopted to improve the IHS in the CIP. The results of the theoretical analysis and NS2 simulation show that packet loss of the real time traffic during the handoff can be reduced dramatically with the EIHS.
Based on the amplify and forward or the decode and forward mode, the channel capacity was analyzed for a multi-antenna technology based two hop cooperative system. Taking the maximum end to end data rate as the optimization criterion, an adaptive power allocation scheme was proposed to distribute a fractional power to each hop according to the channel state information. The simulation results prove the advantages of the cooperative system and show that compared with the uniform power allocation scheme, the proposed power allocation scheme can improve the system capacity by adjusting each hop power adaptively and deciding whether or not to relay based on the channel status.
Based on multi-path fading environment, the impact of simulcasting and signal combining methods on the forward link performance in direct spreading code division multiplex access（DSCDMA）distributed antenna systems with imperfect channel estimation was analyzed. A detailed multi-path multi-access model for distributed antenna systems was proposed and exact outage probability expression was derived. The study shows that the impact of channel estimation errors varies with system load. Furthermore, forward link simulcasting can decrease both the impact of channel estimation errors and the outage probability.
In orthogonal frequency division multiplexing systems, the pilot symbols are often inserted for channel estimation. The frequency offset estimation algorithm using these pilot symbols was further analyzed. By modifying the derivation procedures both in time domain and frequency domain, and by the proper approximation, the approximate maximum likelihood estimation result in time domain was derived. Meanwhile, a different form of likelihood function in frequency domain was derived, and analytical estimation result was also given. The new algorithm decreases the computational complexity greatly, and increases the accuracy of the estimation greatly. The estimation algorithm using pilot symbols together with the decision data was also analyzed, and the accuracy of the estimation result can be improved by utilizing the decision data.
An end to end adaptive multi-protocol label switching （MPLS）based fast handover method for mobile IP（MIP） was proposed to improve the flexibility and scalability of MIP over MPLS. The new method was based on the hierarchical structure of network in operation. Through installing and timely refreshing a mobile node(MN) list in each label switched router （LSR），the deployment of regional foreign agent （RFA） could be optimised dynamically in a distributed way. Theoretical analysis demonstrates that the new method is transparent to the upper layer applications and underlying network services, backward compatible, easy to deploy and cost effective.
To improve the success probability of transmitted packets of leader nodes in tactics Ad Hoc network, a novel backoff algorithm was proposed, which was different from general methods of setting backoff intervals. When the leader succeeded to send a packet, the backoff interval was set as the maximum value. After failing to send a packet, the backoff interval was gradually shortened when times of retransmission increased. Then its performance was theoretically analyzed. Computer simulation shows that the success probability of packets of leader nodes which are transmitted is greatly improved and little effect is given to the success probability of packets of common nodes.
Based on the generalized probability data association (GPDA) algorithm method in quadrature amplitude modulated (QAM) in multi-input multi-output (MIMO) system, the same algorithm was applied to code division multiple access(CDMA) system combined with multiuser detection method, and the softinput softoutput(SISO) GPDA + SISO FEC(forward error correction) joint multiuser detector (MUD) was proposed. The external information was utilized to process interference reducing, decoding and decision. Computer simulation results show this scheme can achieve maximum near optimism performance of single user receiver with less complexity.
An improved crossinterference iterative decoding algorithm for multiple input multiple output(MIMO) structure was proposed to reduce the interference among users employing the same resource in generalized distributed antenna array architecture of beyond 3G(B3G) system. The traditional decoding algorithm for Bell lab layered space time (BLAST) code and interference cancellation by transferring information among multiantenna arrays were combined. The iterative decoding process was performed for several times, which would solve the problem of interference among users in adjacent coverage areas of antenna arrays and improve the performance of the system effectively.
A powercontrolled topologytransparent scheduling algorithm in Ad Hoc networks was proposed. By discovering the number of neighbor nodes periodically, the algorithm could regulate transmitting power to improve the space efficiency and to decrease collisions. The extensive analysis and OPNETbased simulation results show that the proposed algorithm can improve average throughput and decrease transmit block greatly. In addition, the lifetime of the network can be increased by reducing the transmitting power.
The evolving process of channel matrix from traditional joint detection (JD) systems to traditional multiinmultiout (MIMO) systems was presented and a novel generalized MIMO channel matrix model was proposed. The evolving process shows that, MIMO and JD can be unified in this model. Moreover, with such a channel matrix model, the detection algorithm of Bell lab layered space time (BLAST) code in traditional MIMO systems can also be used in JD systems with a better (07?dB) bit error rate (BER) performance.
Timing estimation and multipath delay estimation algorithms for a multi input multi outputorthogonal frequency division multiplexing (MIMOOFDM) system were presented. The basic principle of these algorithms is based on the correlation operation between a local training sequence and the sampled received signal. Noise estimation, 1order infinity impulse response (IIR), and deleting the correlation lobes peak value were introduced to improve the performance of timing estimation. Probability statistics for multipath delay of multiframe was introduced to improve the performance of multipath estimation. Simulation results indicate that the algorithms are robust against a frequency selective fading channel: The timing synchronization algorithm shows satisfactory performance even at a low signaltonoise ratios (SNR), and mean square error (MSE) of 10-7 is obtained at SNR＞6?dB for the multipath delay estimation, while channel sampling rate is 20?MHz.
The hierarchical mobile IP (HMIP) is efficient for reducing signaling cost and handover latency. However, since user mobility characteristics and traffic load between different cells are always changing, the preplaned hierarchical topology is likely to be invalid or even causes more signaling cost. A novel algorithm based on domain discovery and topology adaptive optimization was proposed, which can discover those cells with high handover frequency or heavy traffic load through realtime measurement and make these cells gather into the same administrative domain. In this way, the interdomain handover/traffic will change to intradomain handover/traffic. The theoretic analysis and simulation results both show that the algorithm reduces the signaling costs and handover latency and increases the throughput of mobile nodes obviously.
Baed on decoding performance, complexity and latency of low density parity check (LDPC) codes, a lowcomplexity parallel decoding algorithm was proposed. By simplifying the message sent from check nodes to bit nodes, the complexity of the algorithm approaches that of the “minisum” algorithm. Simulation results show that the performance is very close to belief propagation algorithm in performance.
An adaptive beamforming algorithm based on nonlinear constraint was presented. Firstly, the algorithm modified the eigenvectors corresponding to signal subspace of the covariance matrix of the receiving signals. SOI(signal of interest) was not contained in the modified covariance matrix but interference signal and noise were contained. Then LCMV (linearly constrained minimum covariance) method and the modified covariance matrix were adopted in the new approach to obtain weight vector of adaptive pattern. To improve the robust characteristics, nonlinear constraint method was used to optimize the weight vector of adaptive pattern. The final solution of weight vector was different from alterably diagonal loading and the unknown parameter in optimal solution can be obtained easily and accurately. The SINR (signal to interference plus noise ratio) of the new algorithm has robust characteristics for random steering vector errors and is not sensitive to power change of SOI. Simulation results prove the validity of the new algorithm.
A scheme of extended joint detection was presented, which was used to cancel strong adjacent cell interference (ACI) and intra cell interference. The spread code was judged and the channel characteristic of severe ACI was estimated first. Then the system matrix was extended with such information to joint detect with the users in desired cell, canceling ACI and intra cell interference simultaneously. Simulation results show that the severe ACI in time divisionsynchronous code division multiple access (TDSCDMA) system can be efficiently cancelled, the performance fluctuation can be decreased and the robustness of the system can be enhanced with this scheme.
The dynamic subcarrier allocation algorithm was studied. A novel algorithm for subcarrier allocation called delay weighted dynamic subcarrier assignment (DWDSA) was proposed. DWDSA was composed of inner and outer loop control structure. The function of inner loop was to allocate the subcarrier with the given power allocation scheme. And that of out loop was to adjust the delay weighted parameters. The performance of the queuing delay and buffer occupancy are improved according the simulation results. Moreover, in this algorithm the queuing delay and channel condition were considered to improve the resources utility and to satisfy the requirement of quality of service (QoS).
Based on the singlecarrier frequency domain equalizer（SCFDE）approach providing immunity of intersymbol interference (ISI), a novel frequency domain balanced structure for ultrawideband system was proposed. The optimal frequency domain balance algorithm based on minimum meansquare error (MMSE) estimation wais introduced. The simulation result shows that the bit error ratio can be decreased greatly.
In order to improve the capacity of the 802.16 network, least good channel number first rule was put forward. Based on the rule and 2D mapping characteristics of wireless scheduling in 802.16 system, a wireless resource scheduling algorithm which can ensure maximum delay was proposed. MATLAB simulation and comparison with other algorithms were made. The simulation result indicates that the capacity and the utilization rate of wireless resource in the same wireless channel condition are improved.
Based on the 1Tx and 2Rx systems working in 35?GHz frequency and 20?MHz bandwidth, the typical beyond 3G (B3G) urban wireless propagation environment of China as well as typical indoor channel propagation environment was studied through a recent indoor and outdoor wireless propagation measurement. The general outdoor model of path loss was verified. A new largescale fading mode was proposed, which provided reference model for China B3G channel modeling and system design. Important criteria for B3G system evaluation were provided by the measurement, analysis and statistic results of several parameters, namely 1) r.m.s excess delay and max excess delay in both indoor and outdoor scenarios, 2) the lagtime domain and timefrequency domain channel responses, 3) the correlation between different propagation environment in the singular cell, and the antenna correlation.
A finitestate Markov channel model representing Rayleigh fading channels was introduced. A methodology of selecting signaltonoise thresholds and the total number of states were discussed based on the analysis of the steadystate probability. And the transition probabilities between states and error probabilities were calculated with the given fading characteristic parameters such as level crossing rate. The validity and accuracy of the model are confirmed by the simulation.
The media access control (MAC) mechanisms such as the scheduling and rate control were analyzed. A new combined MAC scheme was proposed based on a comparison of the two approaches. A CDMA2000 1x EVDVRelD system was simulated based on the methodology specified in 3GPP2. The simulation results show that the proposed scheme improves the system stability while increases the throughput and resource utilization of the system. The transmission delay is also reduced so that the quality of service (QoS) requirements of various services can be better satisfied.
The relation between the peak frequency and bandwidth of the nth derivative of the Rayleigh pulse and the pulse shape parameter α and the order of differentiation n was researched. A novel algorithm to design ultrawideband (UWB) pulse was then proposed based on the bandwidth and peak frequency constraints on UWB systems by federal communications commission (FCC). n satisfying FCC mask by schematized version and a range ofα which satisfy the regulation can be figured out by the proposed algorithm, which means a myriad of pulses that are compliant can be designed.
Frequencydomain spectrumoptimalcombination (FDSOC) technology is a novel receiver technology for ultra wideband (UWB) systems. This technology implements the optimal reception of the transmitted data bits in frequency domain by utilizing the fact that the sum of the discrete spectrum components of each UWB pulse is real and the polarity of the sum is variable according to the pulse shape. The principle of FDSOC was analyzed theoretically, and then the performance of FDSOC UWB receiver with frequencydomain subspace channel estimation and UWB Rake receiver with maximum likelihood ( ML) channel estimation were evaluated under UWB channel environment (IEEE 802.15.3a channel model). The simulation results show that FDSOC UWB receiver has good performances.
A kind of predistortion power amplifier by two radio frequency transistors reciprocal compensating was designed. And the temperature compensation circuit was presented. This predistortion power amplifier is verified by many experiments to improve intermodulation distortion(IMD) by 10?dB for a 20?MHz signal bandwidth, and the temperature compensation circuit can maintain IMD change within ±15?dB over the temperature range.
The joint power and data rate allocation in code division multiple access (CDMA) systems was studied. Using utility function as the optimization objective of resource allocation policy, a novel dualloop resource allocation algorithm was presented. Data rate was adjusted by the outer loop to meet fairness among users. Based on the rate allocation achieved by the outer loop, the inner loop allocated power to guarantee the maximum system utility. Outer loop should be controlled by inner loop. The fairness among different users is guaranteed and the system resource utilization is improved with this algorithm.
Outdated neighbor information problems in mobile Ad Hoc network, including outdated neighbor coordinates and outdated neighbor sets, were studied. A localized faulttolerant topology control algorithm MFTTC (mobile faulttolerant topology control) was proposed, which guarantees that any two nodes in the network are both Kvertex ( Svertex) and Kpaths (Spaths) connected under outofdate neighbor information. MFTTC's validity was proved. Simulation results show that by using MFTTC, network topology can be simplified and by varying K,network reliability can be controlled even under outdated neighbor information. Also, the network topology could be adjusted according to the maximum speed of nodes, which is helpful to keep topology stability in a mobile network.
Baed on the way to construct permutation sequence code (PSC) and comparisons of the characteristics of autocorrelation and crosscorrelation, the existence and the low bound of the primitive number in GF(2m) were analyzed. A method on calculating the primitive was proposed. The simulation result shows that PSC code is a kind of efficient code for ultra wide band.
A propagation model calibration method based on propagation environment classification was presented, which took advantages of the information obtained from digital maps. Experiments and simulation results show that the inappropriate application of calibrated propagation models leads to larger errors of path loss prediction, with a mean error around 30?dB. While with the proposed method, the accuracy of path loss prediction could be improved with the mean error decreasing 6?dB to 30dB.
A novel framework was proposed which introduced the Game Theory into the multiantenna power control in broadband wireless systems. Firstly, the power control problem was described as an Nperson noncooperating Game. Secondly, antenna's quality of service（QoS）level was formulated by a utility function and the cost for reaching the level is formulated by a price function. Finally, the process of power control was expressed as the process of maximum net utility (utility function minus price function) for each antenna. For the application of the framework, a new algorithm was also proposed. By the results of theoretical deductions and simulations, the algorithm converges to a fixed Nash equilibrium point and it also has the tunable parameter to meet the needs of different QoS levels.