Large intelligent surfaces with sparse channel sensors
Large intelligent surfaces (LISs) with sparse channel sensors are provided. Embodiments described herein provide efficient solutions for these problems by leveraging tools from compressive sensing and deep learning. Consequently, an LIS architecture based on sparse channel sensors is provided where all LIS elements are passive reconfigurable elements except for a few elements that are active (e.g., connected to baseband). Two solutions are developed that design LIS reflection matrices with negligible training overhead. First, compressive sensing tools are leveraged to construct channels at all the LIS elements from the channels seen only at the active elements. These full channels can then be used to design the LIS reflection matrices with no training overhead. Second, a deep learning-based solution is deployed where the LIS learns how to optimally interact with the incident signal given the channels at the active elements, which represent the current state of the environment and transmitter/receiver locations.
1 . A method for operating a large intelligent surface (LIS), the method comprising:
configuring the LIS with an array of passive reconfigurable reflecting elements and a plurality of active sparse channel sensing elements;
operating the plurality of active sparse channel sensing elements among the array of passive reconfigurable reflecting elements in a channel sensing mode, wherein each of the array of passive reconfigurable reflecting elements has a reconfigurable phase state; and
adjusting the corresponding phase states of the plurality of passive reconfigurable reflecting elements based on output of the channel sensing mode.
2 . The method of claim 1 , further comprising:
operating the plurality of active sparse channel sensing elements in a reflection mode based on the output of the channel sensing mode.
3 . The method of claim 1 , further comprising:
receiving sensor data from at least one of a video sensor, an audio sensor, a radar sensor, or a proximity sensor,
wherein adjusting the plurality of passive reconfigurable reflecting elements is further based on the sensor data.
4 . The method of claim 1 , wherein operating the plurality of active sparse channel sensing elements in the channel sensing mode comprises:
receiving one or more sampled channel vectors; and
determining controllable parameters of the plurality of passive reconfigurable reflecting elements from the one or more sampled channel vectors.
5 . The method of claim 4 , further comprising:
determining the controllable parameters based on extrapolating channels for some or all of the plurality of passive reconfigurable reflecting elements using sampled channels of the plurality of active sparse channel sensing elements.
6 . The method of claim 4 , further comprising:
determining the controllable parameters based on compressive sensing.
7 . The method of claim 4 , further comprising:
determining the controllable parameters based on artificial intelligence.
8 . The method of claim 7 , wherein the artificial intelligence comprises a deep learning algorithm implemented on a neural network and configured to resolve a reflection beamforming vector for the plurality of passive reconfigurable reflecting elements.
9 . The method of claim 8 , further comprising:
pre-training the deep learning algorithm to learn an environment around the LIS.
10 . The method of claim 1 , further comprising:
coupling the LIS to the plurality of active sparse channel sensing elements with one or more radio frequency (RF) chains.
11 . The method of claim 10 , further comprising:
selectively connecting each of the plurality of active sparse channel sensing elements is to a corresponding RF chain,
wherein each of the plurality of active sparse channel sensing elements is operable in a passive reflection mode when disconnected from the corresponding RF chain.
12 . The method of claim 11 , further comprising:
adjusting the plurality of active sparse channel sensing elements when in the passive reflection mode using a reflection matrix.
13 . The method of claim 12 , further comprising:
coupling the LIS to one or more sensors; and
resolving the reflection matrix based on input from the plurality of active sparse channel sensing elements and the one or more sensors.
14 . The method of claim 1 , wherein the array of passive reconfigurable reflecting elements comprises multiple collocated groups of passive reconfigurable reflecting elements.
15 . The method of claim 10 , wherein the array of passive reconfigurable reflecting elements comprises multiple distributed groups of passive reconfigurable reflecting elements.
16 . The method of claim 10 , further comprising:
estimating a wireless channel using the plurality of active sparse channel sensing elements;
adjusting the reconfigurable phase state of each of the array of passive reconfigurable reflecting elements based on the estimated wireless channel; and
resolving a reflection matrix for the reconfigurable phase state of each of the array of passive reconfigurable reflecting elements to facilitate wireless communication over the estimated wireless channel using a deep learning algorithm trained using past inputs from the plurality of active sparse channel sensing elements,
wherein the plurality of active sparse channel sensing elements is fewer in number than the array of passive reconfigurable reflecting elements.
17 . The method of claim 1 , wherein each of the array of passive reconfigurable reflecting elements comprises a corresponding reconfigurable radio frequency (RF) phase shifter.
18 . The method of claim 11 , wherein each of the plurality of active sparse channel sensing elements comprises a corresponding reconfigurable RF phase shifter.
19 . The method of claim 18 , wherein each of the plurality of active sparse channel sensing elements is operable in an active sensing mode and a passive reflection mode.
20 . The method of claim 19 , wherein
in the active sensing mode, each of the plurality of active sparse channel sensing elements is connected to a corresponding RF chain, and
in the passive reflection mode, each of the plurality of active sparse channel sensing elements is connected to the corresponding RF phase shifter and disconnected from the corresponding RF chain.