Analog Least Mean Square Loop for Self-Interference Cancellation: Implementation and Measurements

Abstract

Analog least mean square (ALMS) loop is a simple and efficient adaptive filter to cancel self-interference (SI) in inband full-duplex (IBFD) radios. This paper proposes a practical structure and presents an implementation of the ALMS loop. By employing off-the-shelf components, a prototype of the ALMS loop including two taps is implemented. The prototype is evaluated in IBFD systems which have 20 MHz and 50 MHz bandwidths, respectively, with the carrier frequency of 2.4 GHz. The performance of the prototype with different roll-off factors of the transmit pulse shaping filter is also examined. Experimental results show that the ALMS loop can provide 39 dB and 33 dB of SI cancellation for the two systems, respectively. Furthermore, when the roll-off factor of the pulse shaping filter changes, different levels of cancellation given by the prototype are also demonstrated accordingly. These experimental results validate the theoretical analyses presented in our previous publications on the AT, MS loop behaviors.