Active wireless sensing for rapid information retrieval in sensor networks

Abstract

Most existing information extraction schemes in sensor networks rely on in-network processing that requires information routing and coordination between sensor nodes and incurs a corresponding overhead in delay and energy consumption. In this paper, we propose a viable alternative

Wireless Communication and Sensing in Multipath Environments Using Multiantenna Transceivers

“…Furthermore, it is often desirable to shift the computational burden from the (sensor) nodes to the access point (WIR) for energy efficiency. With these goals in mind, in this section we discuss a novel downlink communication scheme for sending dedicated information to distinct sensors using space-time reversed (STR) versions of their signatures {h k } [65]. The sensor nodes can then retrieve the information intended for them by simply match-filtering to the (common) spread-spectrum waveform, q(t), used in the uplink communication-no channel estimation is required at the sensors.…”

“…(5.147) Sampling x k (t) at the "optimal" sampling time t =T yields the decision statistic, x k , at the k th sensor for detecting the symbol, s k , intended for it [65]: vector of sensor decision statistics, x, with √ N reflecting the beamforming gain, and w ∼ CN (0, I K ) represents the independent noise corrupting the sensor decision statistics. The diagonal entries of √ EN H TH * represent the desired signal terms and the off-diagonal entries represent the interference in x as in (5.148).…”

“…Note that the desired sensor exhibits a significantly higher SNR compared to other sensors due to the STR operation-in general, the richer the multipath, the sharper the ability of STR signaling to focus the signal at the desired sensor (location) while minimizing the interference to other sensors (locations) [29]. Low-complexity linear precoding techniques, analogous to linear MMSE interference suppression in the uplink, can also be used at the WIR to further suppress any residual interference between the different sensor transmissions, thereby improving the reliability of downlink communication [65]. Figure 5.12b shows the average P e of STR signaling with one such interference suppression scheme as a function of per-sensor transmit SNR (E) when K = 108 sensors are simultaneously addressed.…”