This paper introduces new perspectives on analog design space search. To minimize the time-to-market, this endeavor better cast as constraint satisfaction problem than global optimization defined in prior arts. We incorporate model based agents, contrasted with model-free learning, to implement a trust-region strategy. As such, simple feed-forward networks can be trained with supervised learning, where the convergence is relatively trivial. Experiment results demonstrate orders of magnitude improvement on search iterations. Additionally, the unprecedented consideration of PVT conditions are accommodated. On circuits with TSMC 5/6nm process, our method achieve performance surpassing human designers. Furthermore, this framework is in production in industrial settings.

This paper investigates the problem of adaptive distributed beamforming for wireless sensor/relay networks. Existing schemes seek to adaptively achieve distributed phase alignment for beamforming via the help of an error-free, one-bit feedback link from the receiver to all distributed transmitters. Under such setting, we proposed a Deterministic Bisection Search Algorithm (DBSA), designed to achieve distributed phase alignment efficiently in a greedy fashion. Extensive simulations demonstrated significant improvement in the convergence time of the proposed DBSA compared with existing schemes. For wireless sensor/relay networks, inherent issues of system uncertainty such as random node addition or removal often occur due to power depletion or requests from new nodes to join the network. To resolve these issues, DBSA was further extended and numerical experiments have shown that the extended version is indeed robust against such system uncertainty.