PHY-layer link quality indicators for wireless networks using matched-filters

Henry E. Baidoo‐Williams, Octav Chipara, Raghuraman Mudumbai, Soura Dasgupta

Abstract

We present a novel approach to accurate real-time estimation of wireless link quality using simple matched-filtering techniques. Our approach is based on the simple observation that there is a portion of each packet transmission from any given node that does not change from one packet to another; this includes preamble sequences used to synchronize the receiver and also address information in the packet header used for medium access control and routing. Our approach can be thought of as a generalized and simplified variant of standard signal processing techniques that are commonly used for preamble detection, automatic gain control, carrier sensing and other functions in many packet wireless networks. By using a combination of energy detection and correlation techniques, we show that we can effectively detect packet transmissions in real-time with low complexity, without decoding the packets themselves, and indeed, even without detailed knowledge of the packet format. We present extensive experimental results from a software-defined radio testbed to illustrate the effectiveness of this approach for 802.15.4 (Zigbee) networks even in the presence of strong interference signals and low SNR.

Introduction

In wireless networking, accurate Link Quality Indicators (LQI) are essential for efficient routing and medium access control. However, network designers often face a trade-off between simplicity and accuracy. Traditional Received Signal Strength Indicators (RSSI) are computationally inexpensive but "blind"; they measure total energy in the air without distinguishing between legitimate signals, interference, or background noise. Conversely, full packet decoding provides high accuracy but is power-intensive and only yields data for successfully received packets, creating a blind spot for packets lost to collisions.

This paper proposes a "middle-way" sensor: a physical-layer (PHY) link quality estimator that is more sophisticated than a power meter but significantly lighter than a full packet decoder. By leveraging redundant features inherent in wireless protocols, the system can identify specific transmitters and estimate signal strength even under heavy interference.

Methodology

The core of the proposed approach is the exploitation of redundancy. In most wireless protocols, specific segments of a packet—such as the synchronization preamble and header addresses—remain constant for every transmission between a node pair. These segments act as a "fingerprint."

The system employs a three-stage signal detection algorithm at each matched-filter (MF) block to minimize computational overhead:

Algorithm 1: Three-stage signal detection process.
  1. Energy Detection ($T_1$): The receiver first checks if the received signal strength exceeds a threshold $T_1$. This ensures that more complex processing only occurs when the medium is active.
  2. Matched Filter Metric ($T_2$): If energy is detected, the signal is passed through a matched filter—a template of the known redundant symbols. A characteristic peak in the output indicates a potential packet match.
  3. Pearson Correlation Coefficient ($T_3$): A final statistical check compares the received signal to the template. This step is vital for distinguishing the desired signal from non-matching interference.

The researchers explored both protocol-aware modes (using clean preambles) and protocol-blind modes (using noisy recordings of previous transmissions as templates), allowing for link tracking even when protocol specifications are unknown.

Experimental Evaluation

The approach was validated using a software-defined radio (SDR) testbed consisting of Ettus USRP N200 nodes. Standard 802.15.4 (Zigbee) traffic was generated using Crossbow TelosB motes, while pre-recorded GMSK (Gaussian Minimum Shift Keying) packets were injected to simulate a crowded radio environment. This setup allowed for testing the algorithm's robustness against real-world interference and low Signal-to-Noise-plus-Interference Ratios (SNIR).

Key Results

The experimental results demonstrate significant improvements over traditional RSSI. While RSSI plots often show a cluttered mix of all signals in the environment, the matched-filter approach successfully isolates specific links.

Discrimination between two distinct transmitter links.

As shown in the matched filter outputs, the system can clearly distinguish between different transmitters (MF1 vs. MF2). The correlation coefficient effectively "zeroes out" interference from non-matching sources, providing a clean estimate of link quality even when the SNIR is below 0dB—conditions where standard decoders typically fail. Furthermore, the magnitude of the correlation peaks serves as a useful indicator of the current signal-to-interference level.

Conclusion

This research presents a robust, low-complexity method for real-time link quality estimation. By focusing on redundant packet features rather than full payload decoding, the system achieves high discrimination and interference rejection. Future work will focus on optimizing hardware implementations (e.g., on FPGAs) to handle high-density networks and developing automated methods for threshold tuning in dynamic environments.

Summary updated

Abstract

We present a novel approach to accurate real-time estimation of wireless link quality using simple matched-filtering techniques. Our approach is based on the simple observation that there is a portion of each packet transmission from any given node that does not change from one packet to another; this includes preamble sequences used to synchronize the receiver and also address information in the packet header used for medium access control and routing. Our approach can be thought of as a generalized and simplified variant of standard signal processing techniques that are commonly used for preamble detection, automatic gain control, carrier sensing and other functions in many packet wireless networks. By using a combination of energy detection and correlation techniques, we show that we can effectively detect packet transmissions in real-time with low complexity, without decoding the packets themselves, and indeed, even without detailed knowledge of the packet format. We present extensive experimental results from a software-defined radio testbed to illustrate the effectiveness of this approach for 802.15.4 (Zigbee) networks even in the presence of strong interference signals and low SNR.
BibTeX
@misc{baidoowilliams2015phy,
  author = {Henry E. Baidoo‐Williams and Octav Chipara and Raghuraman Mudumbai and Soura Dasgupta},
  title = {{PHY-layer link quality indicators for wireless networks using matched-filters}},
  year = {2015},
  eprint = {1508.04524},
  archivePrefix = {arXiv},
  url = {https://arxiv.org/abs/1508.04524},
}
Figure