Integrated Sensing and Communications for Unsourced Random Access via Spectrum Sharing Compressive Sensing Approach With Massive MIMO Receiver

This paper investigates the problem of uplink multiple access under integrated sensing and communications (ISAC) system in a manner of unsourced random access (URA). In the application scenario of massive machine-type communication (mMTC), recent advance in theory reveals that an ISAC system will be overloaded by unbounded number of active devices with conventional multiple access protocol. An enabling way to support the future massive connectivity is to develop the unsourced ISAC (UNISAC). In this paper, a spectrum sharing compressive sensing-based UNISAC (SSCS-UNISAC) is proposed and offers insight into practical UNISAC multiple access code design. Specifically, the proposed encoder effectively maps the signal components of sensing and communications into nonorthogonal domains of power and code with non-linear projection. By utilizing the diversities in power and code domain, the receiver decodes the signal components iteratively in the spirit of successive interference cancellation. Moreover, the performance analyses on the major decoding procedures of the proposed scheme are elaborated to offer approximations on the potential capacity performance of the proposed scheme. Numerical results demonstrate the viability of the proposed scheme. Compared with conventional access protocol, the well-tuned SSCS-UNISAC produces more than 30.5 dB capacity gain at 1000 user size and the capacity gap between the proposed scheme and the information theory bound is less than 5dB with a massive MIMO receiver.