An independent component analysis (ICA) based method is designed for improving the effect of grid radio monitoring and providing an effective layout of the monitoring system, in which the source signals can be retrieved and sudden emergence of abnormal signal sources can be detected. To analyze the multiple factors affecting the performance of the method, a MATLAB simulation modeling is built. The feasibility and efficiency of the proposed ICA based method are also demonstrated.

A cross layer design of quickly tag reading scheme is proposed，which consists of two steps.At the reservation step，a two-dimensional access which combines slotted-ALOHA and orthogonal code reservation is adopted，and tag direction is estimated by the reservation code and antenna array on the reader.At the tag data reading step，the different spatial characters of the tags are employed to cancel the multiple tag interference(MTI) in the same slot.Simulation results show that tag reading speed is very high and it increases with code length and antenna number.

It is especially important for slotted-ALOHA RFID to estimate tag number in order to allocate slot number optimally and maximize the efficiency The existing tag estimation methods are not accurate enough especially when tags overnumber slots What's more, their applicable ranges are quite limited A novel tag estimation method DRCTE(Derivative Reliabilities Combined Tag Estimation) is proposed Three primary estimations are obtained respectively from the rates of empty, occupy and collide slot number to the whole slot number, then their reliabilities are evaluated from the derivatives of the rates, and at last the three primary estimations are combined to get the final estimation result The proposed DRCTE algorithm not only utilizes the reliable information but also restrains unreliable information, so it is more accurate and its applicable range is larger

The mergence of VoIP and WLAN created a new wireless platform for voice service delivery, VoWLAN. This paper figured out the capacity bottleneck of VoWLAN, and analyzed the problems in current solutions such as M-M. A novel unicast based multiplexing scheme (U-M, Unicast Multiplexing) to improve the capacity and performance of VoWLAN was proposed. Both theoretical and simulation results showed that this scheme improved the voice capacity to nearly 100% comparing to the traditional un-multiplexed scheme, and reduced the packet lost ratio by 10% compared with the M-M scheme. Meanwhile, the U-M scheme is a power-saving scheme in the voice-oriented application as well. Therefore the proposed U-M scheme is an effective solution to increase the voice capacity of WLAN and improve the performance of VoIP service.

According to the feature of ST-OFDM system, we proposed an efficient CFO estimation algorithm, which focus on the IFO estimation algorithm. Compared to the conventional algorithm, this novel algorithm can solve the problem that the wide acquisition range/low probability of failure conflicts with low complexity. The simulation results indicate under multi-path channel for low SNR, the new algorithm has lower probability of failure than the conventional algorithm by adopting proper training symbols. When the probability of failure is needed to be lower and the acquisition range is needed to be wider, the gain of proposed algorithm will be larger.