Hybrid TOA/AOA NLOS Identifying And The Research Of NLOS Mitigation

Cellular networks positioning has drawing more and more attention along with the rapid development of wireless communication technical. Different from Globe Positioning System (GPS) which is dedicated to provide location information, positioning in the cellular network is a value-added service. Cellular networks location contains two steppes:step one is extracting distance information from communication signals, such as time of arrival(TOA)、time difference of arrival(TDOA), angle of arrival(AOA)and receive signal strength(RSS).Step two is using different location algorithm to locate the mobile station (MS).Non-line-of-sight (NLOS) transmission is the major aspect that influences the degree of location accuracy. There are two methods to deal with this problem:one is using optimization algorithms to reduce the influence, the other is identifying and discarding NLOS, locating the MS only by line-of-sight(LOS) transmission. To be aimed at these two methods of treatment of NLOS, this paper proposes the three step linear least squares algorithm of low complexity and a hybrid adaptive NLOS identification algorithm respectively.Linear least square (LLS) algorithm is one of the NLOS mitigation algorithms with lower computation. This algorithm chooses a base station (BS) as the referred point at first, and then uses an alternative method to convert the nonlinear distances between the MS and the BSs to be linear. The MS’position is achieved by employing least square (LS) algorithm. This paper proposes a three steppes LLS (TLLS) algorithm:Step one is taking every BS as the referred point and calculates MS’position. Step two is linearizing the distances using Taylor series expansion. After that, the positioning error can be achieved. Step three is taking the MS with the minimum positioning error as the ultimate MS’location. The simulation result shows that TLLS is superior to the conventional algorithms.Although the NLOS mitigation algorithm reduces the influence of NLOS transmission to the location process, but it can’t resolve the dilemma that the positioning error is huge when using the NLOS to locate MS. This paper proposes a method to identify and discard NLOS which combines TOAs and AOAs. It depends on the following:one is that measured AOAs are almost accurate in LOS. Two is that measured AOAs are deviate from the true value in NLOS. Based on the two points, I raise a ergodic method that takes any two or more BSs to get a variety of combinations. Then the paper calculates MS and the AOAs between MS and the BSs. After that, the standard deviations of AOAs are also achieved. Taking the combination which makes the standard deviation minimum as the LOS combination, the MS’location for this combination is the ultimate location. In addition, the measured distances of the BSs which are not belonging to the LOS combination can be converted to a linear constraint that can be used to prove the reliability of the LOS combination. The novel algorithm can attain accurate MS’ location when there are only two LOS BSs. The simulation result presents the proposed algorithm can achieve high accuracy and identify NLOS with high probability.