Warren J. Gross

Enhancing the efficiency of deep neural networks enables complex computer vision applications, such as semantic segmentation, to meet latency requirements on resource-limited embedded systems. By optimizing network parameters, computation overhead can be reduced, leading to lower inference latency. We present TinySegAdapt, a zero-shot architecture search framework for deploying semantic segmentation networks on the ultra-low-power TinyML device. TinySegAdapt compresses networks through parameter-tuning-based architecture search, achieving 2.75 percentage points higher mIoU and 34.2% lower latency compared with filter pruning method. We perform accuracy measurement using a pretrained network and weight sharing, rather than accuracy proxies, yielding a 1.30 pp mIoU improvement. Moreover, TinySegAdapt accelerates the search process by 81.3% through evaluating models on a subset of the validation set without fine-tuning. TinySegAdapt finds competitive designs in just 36.5 hours to discover a variant that is 2.2× faster network on TinyML hardware with a 1.49 percentage point mIoU drop on the PASCAL VOC 2012 dataset.

To extend the applications of polar codes within next-generation wireless communication systems, it is essential to incorporate support for Incremental Redundancy (IR) Hybrid Automatic Repeat Request (HARQ) schemes. For very high-throughput applications, Successive Cancellation (SC) decoding is particularly appealing for polar codes owing to its high area efficiency. In this paper, we propose modifications to SC decoders that employ special nodes to accelerate decoding. Our modifications enable the use of polar IR-HARQ with SC decoding for high throughput applications. Compared to the unmodified SC IR-HARQ scheme, our proposed approach allows us to achieve up to 66% reduction in node traversals with polar codes. Simulation results confirm that the proposed special node modifications do not cause any degradation in FER performance compared to the unmodified state-of-the art special nodes.

The growing adoption of self-supervised learning transformers for speech (speech SSL) is constrained by their significant computational and memory demands, making deployment on resource-constrained edge devices challenging. We propose a latency-aware compression framework that integrates structured pruning and quantization to address these challenges. Guided by a latency model that considers the combined effects of pruning and quantization, our method dynamically identifies and removes less critical blocks while maintaining task performance, avoiding the inefficiencies of over-pruning and under-pruning seen in prior approaches. Unlike prior methods specialized in either post-training compression without fine-tuning data or in cases where fine-tuning data is available, our method is effective in both settings. Experimental results show that, in task-agnostic compression, our method achieves a 4.2× speedup on the Hikey970 edge development platform, outperforming previous task-agnostic pruning methods in most tasks, while requiring only 21–24 GPU hours—a 3× reduction compared to prior methods. Additionally, our method achieves a lower word error rate of 7.8% using task-specific pruning, while reducing computational overhead by approximately 19.4% in terms of GFLOPs compared to previous task-specific methods. Finally, our method consistently achieves higher accuracy than the state-of-the-art post-training compression approach across various latency speedup constraints, even without fine-tuning data.

Implementing automorphisms requires hardware for routing, and it leads to a large area overhead, especially when the number of automorphisms is large.We propose methods for selecting automorphisms in successive cancellation (AED-SC) decoders to reduce the number of routes. We established the equivalence between the automorphism selection and the NP-hard minimum K-union (MKU) problem. To maintain decoding performance under the selected automorphisms, the equivalent class property of the AED-SC is formulated as a quadratic constraint. Also, by incorporating the Hamming distance-based selection method, the problem size of the MKU problem with the quadratic constraint is reduced while enhancing decoding performance. Two heuristics are proposed: a sequential algorithm based on the least expanding set problem, and a greedy algorithm with a complexity that grows linearly with the problem size. Up to a 65% reduction in the number of routes is returned by proposed methods when compared to the state-of-the-art result, while the returned automorphism sets yield comparable decoding performance to the randomly selected set. Synthesis is performed for the routing unit for polar codes with a code length n = 128, a dimension k = 60, and 63 automorphisms that require routing, and these synthesis results verify that a 59% reduction in the number of routes reduces 57% of the logic area required by the routes, which leads to an overall area reduction of 25%.

Targeting sixth generation (6 G) extreme connectivity, massive multiple-input multiple-output (mMIMO) systems face challenges in achieving both high data rates and low latency. Polar turbo product codes (polar-TPCs) with near-optimal error correction performance under constrained latency are promising for 6 G mMIMO applications. In this letter, we propose a dual-stage list (DSL) decoding scheme for polar-TPC coded mMIMO systems, leveraging a turbo list decoder followed by a tree-search-based post-processing method to enhance receiver performance. Moreover, we introduce a two-phase codeword validity check strategy to address the complexity challenges. Empirical results demonstrate that the receiver with our proposed DSL decoding outperforms existing receivers, achieving a superior trade-off between performance and complexity.

Short polar-like codes are competitive for low latency requirements in future communications. Systematic polar codes have not been shown to offer substantial benefits for decoders beyond improving the bit error rate. In this paper, we demonstrate that the sparsity of the equivalent generator matrix of systematic polar codes significantly reduces calculation complexity when using sphere decoding (SD). We propose a fast SD (Fast-SD) for systematic polar codes. Numerical results indicate that the proposed Fast-SD reduces calculation complexity by up to 33.25% compared to SD on short high-rate codes while maintaining maximum likelihood performance.

To extend the applications of polar codes within next-generation wireless communication systems, it is essential to incorporate support for Incremental Redundancy (IR) Hybrid Automatic Repeat Request (HARQ) schemes. For very high-throughput applications, Successive Cancellation (SC) decoding is particularly appealing for polar codes owing to its high area efficiency. In this paper, we propose modifications to SC decoders that employ special nodes to accelerate decoding. Our modifications enable the use of polar IR-HARQ with SC decoding for high throughput applications. Compared to the unmodified SC IR-HARQ scheme, our proposed approach allows us to achieve a 72% reduction in node traversals with a polar code of length 2048. Simulation results confirm that the proposed special node modifications do not cause any degradation in FER performance.