Quadratic Assignment Problems are a class of NP-hard combinatorial optimization problems with a wide range of real-world applications such as Vehicle Routing and FPGA Block Placement. Despite technological advances, solvers that target Quadratic Assignment Problems still require significant computing resources and time, especially as problem sizes grow; with the end of Dennard Scaling leading to the increased development of domain-specific hardware for such tasks. This paper presents MAQO: a hardware architecture for a scalable many-core annealer for quadratic optimization, implemented on an Intel Stratix 10 FPGA. MAQO is comprised of 32 domain-specific processing cores, supporting Quadratic Assignment Problems with up to 200 integer variables, in combination with a Parallel Tempering controller, all operating at 220MHz with a maximum FPGA power draw of 40W. We benchmark MAQO's performance, via solving some of the most difficult known Quadratic Assignment Problem instances, and show that it is, on average, 2.5 times faster than the next best competing solver with 22 times better performance-per-watt.

Boltzmann Machines are recurrent neural networks that have been used extensively in combinatorial optimization due to their simplicity and ease of parallelization. This paper introduces the Permutational Boltzmann Machine, a neural network capable of solving permutation optimization problems. We implement this network in combination with a Parallel Tempering algorithm with varying degrees of parallelism ranging from a single-thread variant to a multi-threaded system using a 64-core CPU with SIMD instructions. We benchmark the performance of this new system on Quadratic Assignment Problems, using some of the most difficult known instances, and show that our parallel system performs in excess of 100 $$\times $$ faster than any known dedicated solver, including those implemented on CPU clusters, GPUs, and FPGAs.

This paper presents an ultra-low power wireless transceiver specialized for but not limited to medical implantable applications. It operates at the 402-405-MHz medical implant communication service band, and also supports the 420-450-MHz industrial, scientific, and medical band. Being IEEE 802.15.6 standard compliant with additional proprietary modes, this highly configurable transceiver achieves date rates from 11 kb/s to 4.5 Mb/s, which covers the requirements of conventional implantable applications. The phase-locked loop-based transmitter architecture is adopted to support various modulation schemes with limited power budget. The zero-IF receiver has programmable gain and bandwidth to accommodate different operation modes. Fabricated in 40-nm CMOS technology with 1-V supply, this transceiver only consumes 1.78 mW for transmission and 1.49 mW for reception. The ultra-low power consumption together with the 15.6-compliant performance in term of modulation accuracy, sensitivity, and interference robustness make this transceiver competent for various implantable applications.

The wireless communication channel around/in the human body is a difficult propagation environment. This paper presents measurement and simulation results to characterize such a channel. A fluid human body model is employed to emulate the inside of a human body. The paper details the fluid human model and path loss model parameters at 400 MHz (MICS band). It is shown that the simulated and measured results are in a close agreement, for instance at a distance of 20 cm and a implant depth of 10 cm, the measurement results in a path loss of -42.1 dB and the simulation in -43.0 dB. The effect of human model shape on measured path loss is analyzed. Furthermore, simulations are employed to characterize this effect. Using the path loss model a top-level link budget is evaluated to determine the feasibility of a given implant device compliant to IEEE802.15.6-WBAN-400 MHz standard.