TR2018-167
Packet Separation in Phase Noise Impaired Random Access Channel
-
- "Packet Separation in Phase Noise Impaired Random Access Channel", IEEE Global Communications Conference (GLOBECOM), DOI: 10.1109/GLOCOM.2018.8647403, December 2018.BibTeX TR2018-167 PDF
- @inproceedings{Pajovic2018dec,
- author = {Pajovic, Milutin and Ozcan, Gozde and Koike-Akino, Toshiaki and Wang, Pu and Orlik, Philip V.},
- title = {Packet Separation in Phase Noise Impaired Random Access Channel},
- booktitle = {IEEE Global Communications Conference (GLOBECOM)},
- year = 2018,
- month = dec,
- doi = {10.1109/GLOCOM.2018.8647403},
- url = {https://www.merl.com/publications/TR2018-167}
- }
,
- "Packet Separation in Phase Noise Impaired Random Access Channel", IEEE Global Communications Conference (GLOBECOM), DOI: 10.1109/GLOCOM.2018.8647403, December 2018.
-
MERL Contacts:
-
Research Areas:
Abstract:
A growing number of applications may benefit from embedding receiver with the capability to separate collided packets directly on a physical layer. For example, in IoT’s random access channel scenario, a number of IoT devices are placed in the same cell, occasionally transmit messages, and are assigned non-orthogonal channel resources, giving rise to packet collisions. While channel access techniques for sharing a common channel are usually employed in most multi-user communication systems, they cause channel underutilization and increased latency. This paper considers a scenario where multiple users asynchronously transmit packets over a shared channel and are not subject to a random access mechanism, or the mechanism itself fails to prevent packet collisions. Assuming that each packet consists of a common preamble and information bearing payload, and experiences random delay, frequency offset, phase noise variation and block flat fading channel, we propose a packet separation algorithm which recovers collided packets, estimates their parameters and detects corresponding payload symbols. The performance of the proposed method is validated using simulations and benchmarked against bounds.