Three researchers from the INÜ presented their work at the IEEE Global Communications Conference (Globecom) held in Taipei, Taiwan from 8-12 December 2025. The Globecom is one of the flagship conferences of the IEEE Communications Society with over 2000 participants from both academia and industry and offers a wide range of technical symposia, keynotes, workshops, exhibitions, tutorials and panels around the topic of future communication systems.
The INÜ contributed the following papers:
- Round-Trip Time Analysis and Optimization for Multi-Link Wireless LANs
- Long Polar vs. LDPC Codes under Complexity-Constrained Decoding
- ORCAS Codes: A Flexible Generalization of Polar Codes with Low-Complexity Decoding
Contribution to Multi-Link WLAN Systems:
Low round-trip times (RTTs) are required for applications like VR/AR and machine control. In our work, we compare the RTTs of multiple state-of-the-art methods in wireless LANs using MAC layer simulations. Moreover, we propose a novel multi-link subband full-duplex operation that is shown to have superior RTT compared to Wi-Fi 7 (IEEE 802.11be).
Long Polar vs. LDPC Codes:
We challenge the common belief that polar codes lose competitiveness to LDPC codes at large block lengths by accounting for realistic complexity constraints on decoding. We show that, for a comparable number of fixed-point LLR operations, long polar codes decoded with SC – and especially SSC – outperform LDPC codes, with SSC achieving sub-logarithmic complexity scaling (per information bit) and requiring fewer operations than even a single BP iteration of an equivalent LDPC code.
ORCAS Codes:
We construct optimally recursively concatenated simplex (ORCAS) codes by Plotkin concatenating near-optimal low- and high-rate simplex codes and their duals, which admit low-complexity maximum likelihood decoders. For practical parameters, ORCAS codes outperform polar codes in block error rate by up to 0.5 dB under successive cancellation decoding, while maintaining similar complexity and offering more flexible codeword lengths.