[Paper Review] Measurement Based Interference Models for Wireless Scheduling Algorithms.

Helga Gudmundsdottir, Eyjólfur Ingi Ásgeirsson+6 more|arXiv (Cornell University)|Jan 8, 2014

Energy Efficient Wireless Sensor Networks46 references8 citations

TL;DR

This paper proposes a measurement-based interference model for wireless scheduling that replaces idealized distance-based path loss with a real-world gain matrix derived from RSS measurements. The model improves prediction accuracy over traditional models and introduces a parameter ζ to quantify environmental complexity, enabling theory-practice alignment in wireless network design.

ABSTRACT

Modeling physical layer behavior of packet reception in the presence of interference is central to achieving efficient spectrum use in wireless sensor networks via spatial reuse. On one hand, analytic and simulations research has largely relied on assumptions of geometric path loss and isotropic transmission which have not been borne out in experiments. Experimental research, on the other hand, has not adopted theoretical models and instead focused on measuring the reality on the ground. We propose a new framework for wireless algorithms. First, distance-based path loss is replaced by an arbitrary gain matrix, typically obtained by measurements of received signal strength (RSS). This allows for the modeling of complex environments, e.g., with obstacles and walls. Second, a new parameter ζ indicates how close the gain matrix is to a distance metric, effectively measuring the complexity of the environment. We experimentally validate our framework on two indoors testbeds with 20 and 60 motes. The results validate the basic properties of the model, the predictive ability of packet reception, dominance over distance-based models, and the sensitivity of ζ to the nature of the environment. Theoretically, we show that all known SINR scheduling algorithms that work in general metric spaces carry over and achieve equivalent performance guarantees in the new model. The conclusions suggest that wireless theory can finally be grounded in experimental practice.

Motivation & Objective

• Address the gap between theoretical wireless scheduling models and real-world physical layer behavior in complex indoor environments.
• Overcome limitations of traditional models based on geometric path loss and isotropic transmission, which do not reflect real-world propagation effects.
• Develop a practical, experimentally grounded framework that enables accurate prediction of packet reception under interference.
• Introduce a parameter ζ to quantitatively assess how far the actual propagation environment deviates from a simple distance-based model.
• Demonstrate that existing SINR-based scheduling algorithms can be applied in the new model with equivalent theoretical performance guarantees.

Proposed method

• Replace distance-based path loss with an arbitrary gain matrix derived from empirical measurements of received signal strength (RSS) across multiple transmitter-receiver pairs.
• Use multilateration or calibration techniques to construct the gain matrix from real-world RSS data collected in indoor testbeds.
• Introduce the parameter ζ as a metric to quantify the deviation of the actual channel gain from a pure distance-based model, reflecting environmental complexity.
• Validate the model’s predictive power by comparing predicted packet reception rates against actual measurements in two indoor testbeds with 20 and 60 motes.
• Prove theoretically that all known SINR scheduling algorithms designed for general metric spaces remain valid and achieve equivalent performance guarantees in the new model.
• Apply the model to evaluate and compare the performance of scheduling algorithms under real propagation conditions, including multipath and blockage effects.

Experimental results

Research questions

• RQ1 To what extent does a measurement-based gain matrix improve the accuracy of interference modeling compared to idealized distance-based path loss models?
• RQ2 How well can the proposed model predict actual packet reception outcomes in real indoor wireless environments?
• RQ3 How does the parameter ζ reflect the physical complexity of the environment, such as obstacles and multipath effects?
• RQ4 Can existing theoretical scheduling algorithms designed for metric spaces be adapted to work under the new measurement-based model with equivalent performance guarantees?
• RQ5 What is the practical impact of environmental complexity (as captured by ζ) on spatial reuse and interference management?

Key findings

• The measurement-based gain matrix significantly outperforms traditional distance-based models in predicting packet reception, especially in non-line-of-sight and obstructed environments.
• The model demonstrates strong predictive accuracy in both indoor testbeds with 20 and 60 motes, confirming its robustness across different network scales.
• The parameter ζ effectively captures environmental complexity, showing higher values in cluttered environments with walls and obstacles, and lower values in open or free-space conditions.
• The model enables the application of existing SINR-based scheduling algorithms in real-world settings with equivalent theoretical performance guarantees, bridging theory and practice.
• Experimental results confirm that the model’s predictions align closely with observed interference patterns, validating its use for practical wireless scheduling design.
• The framework enables more efficient spatial reuse by accurately modeling interference, leading to improved spectrum utilization in real deployments.