Three Level Recognition Based on the Average of the Phase Differences in Physical Wireless Parameter Conversion Sensor Networks and Its Effect to Localization with RSSI

Abstract

In recent years, there have been increased demands for aggregating sensor information from several sensors owing to the spread of the Internet of Things (IoT). However, packet communication, which is a conventional multiple-access technology, is hindered by packet collisions owing to simultaneous access by sensors and waiting time to avoid packet collisions; this increases the aggregation time. The physical wireless parameter conversion sensor network (PhyC-SN) method, which transmits sensor information corresponding to the carrier wave frequency, facilitates the bulk collection of sensor information, thereby reducing the communication time and achieving a high aggregation success rate. However, when more than one sensor transmits the same frequency simultaneously, the estimation accuracy of the number of accessed sensors deteriorates significantly because of multipath fading. Thus, this study focuses on the phase fluctuation of the received signal caused by the frequency offset inherent to the sensor terminals. Consequently, a new feature for detecting collisions is proposed, which is a case in which two or more sensors transmit simultaneously. Furthermore, a method to identify the existence of 0, 1, 2, or more sensors is established. In addition, we demonstrate the effectiveness of PhyC-SNs in estimating the location of radio transmission sources by utilizing three patterns of 0, 1, and 2 or more transmitting sensors.

Enhanced sensitivity of collision detection by carrier frequency offset switching in physical wireless parameter conversion sensor networks

“…When multiple sensors transmit the same subcarrier number, collisions occur at the aggregation station. A previous study [2] noted that the phase of the received signal at the time of subcarrier collision fluctuates for each received antenna because the frequency offset, i.e., the frequency difference in the local oscillator between the transmitter and receiver, differs with each sensor. Consequently, using this phase variation, collision detection was realized via the simultaneous access of multiple sensors.…”

“…To suppress the effect of intercarrier interference caused by the collapse of orthogonality owing to the frequency offset, a long-interval FFT applying the Blackman-Harris window was proposed [2]. When the signal time length increases, the periodicity of the transmitted subcarrier wave increases, the main lobe of the frequency spectrum becomes sharper, and the side lobes remain low, thus suppressing intercarrier interference.…”

“…In Figure 1, the number of signal transmissions was set to two for simplicity; however, it can be expanded to an arbitrary number. First, the aggregation station sends a control signal to all the sensors requesting aggregation, as in the case of Ref [2]. Following the signal, each sensor transmits a subcarrier wave based on the detected sensing data.…”

“…Essentially, they send signals with carrier frequencies switched for sensing data and detect the sensor information reported by the sensors that are accessing simultaneously with the frequency power distribution of the received signal [1]. A previous study reported a method for detecting collisions that occur when multiple sensors transmit the same sensing data in a PhyC-SN; the method employs the average phase variance between the receiving antennas as an identifier [2]. Collision detection is necessary to recognize two or more sensors that form the same data as the aggregation station.…”

Indoor Localization System Based on RSSI-APIT Algorithm

“…To preserve the integrity of data collection, it is necessary to eliminate outliers in this particular segment. Residual plots, when used in conjunction with linear regression, are commonly employed to identify outliers [ 16 ]. In order to conduct linear regression, it is necessary to first identify and exclude singular RSSI values that significantly differ from the normal range.…”