Mitigating cellular network congestion through adaptive beamforming with reflectarrays
Growing congestion in next-generation cellular networks such as 5G is driving the use of previously unused portions of the electromagnetic spectrum, such as MMW frequencies. As with any radio-frequency transmissions, radio receivers in dense MMW cellular networks need to desensitize toward unwanted signals while preserving desired signals. One technique to do so is adaptive beamforming, which is traditionally based on an array of spaced-apart antennas that capture RF signals (desired and interference) from various directions; individual receivers per antenna convert each signal into digital form, and a digital signal processor dynamically updates the signal weight (phase and amplitude) in real time as the signal environment changes. The processor then combines the weighted signals to focus the beam in the intended signal direction. However, the size, weight, power, and cost requirements of many-element fully digitized designs are driving the development of alternative approaches to adaptive beamforming.
A team at Lincoln Laboratory is developing one such alternative, which requires only one or two digital receivers and uses a large reflectarray antenna panel containing hundreds of digitally reconfigurable reflectors. Incoming analog signals reflect off these reflectors, and one stream of samples (the total of the reflections) is collected through a feed horn (funnel-like component that guides signals from the antenna toward the receiver) facing the panel, before the receiver digitizes them. The challenge was to design a panel state that isolates the wanted signals in the sample stream. The team developed iterative beamforming algorithms that work with the limited information available as a result of the single digitization point. With these algorithms, they successfully controlled the scanning beams of a Laboratory-developed prototype reflectarray antenna to reject unwanted transmissions. In addition to this hardware experiment, they simulated the prototype reflectarray and successfully ran scenarios in which the algorithms simultaneously rejected interference sources. Future work includes refining the algorithms to explicitly preserve intended signals, and evaluating realistic effects such as users in motion.
