Packet Detector Using LFSR Seed Cross-Correlation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional packet detection methods in multi-user communication systems using OFDM transmission suffer from performance loss due to threshold-based cross-correlation approaches, particularly in environments with neighboring networks, resulting in up to 3 dB of performance degradation.
Innovation Solution
Implementing receiver algorithms that utilize knowledge of the seed of the constellation scrambler LFSR used by neighboring networks, comparing detection results using local and out-of-domain LFSR seeds to accurately determine packet origin, and communicating this information through packet headers or MAP frames to differentiate between in-domain and out-of-domain packets.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Device complexity
If threshold-based cross-correlation approach is used for packet detection, then packet detection can be performed with simple algorithms, but detection accuracy deteriorates by up to 3 dB in environments with neighboring networks
Solution Approach 1:
The patent segments the packet detection process into two distinct phases: first performing cross-correlation using a local LFSR seed to detect in-domain packets, then performing additional cross-correlation using an out-of-domain LFSR seed to detect and identify neighboring network packets. This segmentation allows the system to maintain simple detection algorithms while improving accuracy by separately handling different packet sources.
Solution Approach 2:
The patent applies preliminary action by performing cross-correlation with the local LFSR seed first to identify in-domain packets before attempting to detect out-of-domain packets. This preliminary detection step allows the receiver to prioritize processing of relevant local network traffic while still maintaining capability to identify and filter neighboring network interference.
2Loss of information
If multiple LFSR seeds are used to differentiate in-domain and out-of-domain packets, then packet origin identification improves, but device complexity increases
Solution Approach 1:
The patent changes the parameter of LFSR seed values to differentiate between in-domain and out-of-domain packets. By using distinct seed values for local and neighboring networks, the system can identify packet origins through cross-correlation without requiring complex identification mechanisms. The receiver simply compares correlation results using different seed values to determine packet source.
Solution Approach 2:
The patent uses copying by generating predefined signals that are copies of the expected packet structures modulated with different LFSR seeds. The receiver creates local copies of what in-domain and out-of-domain packets should look like, then compares received signals against these copies through cross-correlation to identify packet origins efficiently.
Data Source
AI summary
There are several exemplary ways to more efficiently communicate an out-of-domain seed to a receiver—in a first technique, the seed can be indicated in the header portion or data portion of a packet. For example, the header portion of the packet could contain one or more bit fields that indicate the value of the LFSR seed used for the preamble portion of the packet. The receiver would learn the out-of-domain seed after receiving a first out-of-domain packet and decoding the header portion of that packet. After learning the out-of-domain seed, the receiver could send a packet indicating the value of the out-of-domain seed to the local master. The local master could then transmit the value of the out-of-domain seed in the header portion or data portion of a local MAP frame.