Protocol Design for Irregular Repetition Slotted ALOHA With Energy Harvesting to Maintain Information Freshness

Basic Information

• Paper ID: 2411.01446
• Title: Protocol Design for Irregular Repetition Slotted ALOHA With Energy Harvesting to Maintain Information Freshness
• Authors: Khac-Hoang Ngo (Linköping University), Diep N. Nguyen (University of Technology Sydney), Thai-Mai Dinh Thi (VNU University of Engineering and Technology)
• Classification: cs.IT (Computer Science - Information Theory), math.IT (Mathematics - Information Theory)
• Publication Date: arXiv preprint, submitted November 2024, updated January 2, 2025
• Paper Link: https://arxiv.org/abs/2411.01446v2

Abstract

This paper investigates an Internet of Things (IoT) system where energy harvesting devices transmit status updates to a common receiver using the Irregular Repetition Slotted ALOHA (IRSA) protocol. Energy shortages on devices may cause transmission failures unknown to the receiver, disrupting the decoding process. To address this issue, the authors propose a method enabling the receiver to perfectly identify such failures. Furthermore, the information freshness of status updates is enhanced by optimizing the degree distribution of the protocol. The optimized degree distribution mitigates the adverse effects of potential transmission failures. Numerical results demonstrate that despite energy harvesting constraints, IRSA can achieve information freshness levels comparable to systems with unlimited energy.

Research Background and Motivation

Problem Definition

1. Core Problem: In IoT systems, transmission failures caused by insufficient energy when energy harvesting devices use the IRSA protocol to transmit status updates affect the receiver's decoding process and information freshness.
2. Significance:
   • IoT devices are typically deployed in remote locations where battery replacement is impractical
   • Energy harvesting is a key solution for achieving long-term low-power operation
   • Time-critical applications require guaranteed information freshness
3. Limitations of Existing Approaches:
   • Traditional IRSA assumes all intended copies are successfully transmitted
   • Transmission failures caused by energy harvesting occur at unknown locations to the receiver, disrupting the Successive Interference Cancellation (SIC) process
   • Existing research assumes the receiver knows the locations of discarded copies but does not explain how to achieve this
4. Research Motivation: Design an IRSA protocol capable of handling unknown transmission failures and optimize the degree distribution to maintain information freshness

Core Contributions

1. Proposes a method for the receiver to perfectly identify transmission failures: Based on the conventional assumption that the receiver can identify collision-free slots without requiring additional information
2. Validates critical assumptions: Demonstrates the feasibility of the key assumption in previous research that "the receiver knows the locations of discarded copies"
3. Protocol optimization: Optimizes the degree distribution of IRSA to minimize average Age of Information (AoI)
4. Performance analysis: Provides theoretical analysis of lower bounds on packet loss rate under energy harvesting constraints
5. Experimental validation: Demonstrates that optimized IRSA can achieve information freshness approaching systems with unlimited energy despite energy harvesting constraints

Methodology Details

Task Definition

• Input: U energy harvesting devices generating timestamped status updates
• Output: Receiver successfully decodes status updates while maintaining information freshness
• Constraints:
  • Limited battery capacity E
  • Stochastic energy harvesting process (collecting 1 energy unit per slot with probability η)
  • Collision channel model (multi-packet collisions cause decoding failure)

System Model

Energy Harvesting Model

• Battery capacity: E energy units
• Energy harvesting: Independently collect 1 energy unit per slot with probability η
• Transmission cost: Each packet transmission consumes 1 energy unit
• Energy collection stops when battery is full

IRSA Protocol

• Time divided into frames of M slots
• Active devices send L identical copies to L randomly selected slots
• Degree L follows probability distribution {Λℓ}, denoted as Λ(x) = Σℓ Λℓxℓ
• Receiver uses SIC decoding

Technical Innovations

1. AVOID Scheme

Devices only use available energy at frame start for transmission, i.e., Λℓ,b = 0 for ℓ > b, ensuring all intended copies can be transmitted.

Initial Battery Level Evolution (Theorem 2):

P[Bj+1 = b2 | Bj = b1] = Σℓ Ξℓ,b1 × Bino(b2-b1+ℓ; M, η)

2. IDENTIFY Scheme

Allows devices to plan more copies than initial battery capacity, identifying discarded copies through the following algorithm:

Algorithm Flow:

1. Maintain candidate list Sn for each slot n
2. Find singleton slots, decode packets, and add to respective candidate lists
3. For each slot, attempt to remove all possible subsets of candidate lists
4. If removing a subset yields a singleton slot, all packets in that subset have been transmitted
5. Repeat until no more singleton slots are found

Performance Guarantee (Theorem 3): Under infinite SIC iterations, the IDENTIFY scheme achieves the same packet loss rate as when the receiver knows the locations of discarded copies.

3. Packet Loss Rate Lower Bound

Theorem 1: The lower bound on steady-state packet loss rate is:

Pe ≥ φ0[Σy=1^M η(1-η)^(y-1) Σℓ=0^ℓmax Λℓ,0 × (y-1)!(M-ℓ)!/((y-ℓ-1)!M!) + (1-η)^M]

Protocol Optimization

Optimization objective: Minimize average AoI

minimize Δ̄ = 1/α + M(3/2 + 1/ξ - 1/σ)
subject to: Λℓ,b ∈ [0,1], Σℓ Λℓ,b = 1

where ξ = σ(1-Pe) is the AoI reset probability.

Experimental Setup

Parameter Configuration

• Number of devices: U = 1000
• Frame length: M = 100 slots
• Battery capacity: E = 2 energy units
• Energy harvesting rate: η = 0.02 units/slot
• Maximum degree: ℓmax = 5

Evaluation Metrics

1. Packet Loss Rate (PLR): Probability that transmitted packets are not successfully decoded
2. Throughput: G(1-Pe) packets/slot
3. Average Age of Information (AoI): Information freshness at the receiver regarding the tracked process
4. Age Violation Probability (AVP): Probability that AoI exceeds threshold θ

Comparison Methods

1. Slotted ALOHA: Devices immediately transmit generated updates
2. Unlimited Energy IRSA: Serves as ideal benchmark
3. AVOID vs IDENTIFY: Two schemes for handling energy constraints

Implementation Details

• Nelder-Mead algorithm for degree distribution optimization
• Monte Carlo simulation over 10^5 frames
• Multiple random initializations with best result selection

Experimental Results

Main Results

Packet Loss Rate and Throughput

• Low channel load: PLR of AVOID and IDENTIFY significantly higher than unlimited energy case
• High channel load: PLR of both schemes slightly lower due to reduced collisions from packet discarding
• IDENTIFY overall outperforms AVOID

Information Age Performance

• IDENTIFY vs AVOID: At αU=1, ηM=4, AVOID's average AoI is 24% higher than IDENTIFY
• Comparison with unlimited energy: Optimized IRSA only slightly exceeds unlimited energy system
• Comparison with slotted ALOHA: IRSA's average AoI reduced by 40.4%

Parameter Impact Analysis

1. Update frequency α: AoI monotonically decreases with increasing α
2. Battery capacity E: Larger capacity supports higher degrees, improving performance
3. Energy harvesting rate η: Higher collection rate reduces energy shortage
4. Frame length M: Optimal value exists balancing collection opportunities and transmission delay

Key Findings

1. IDENTIFY Advantages Conditions:
   • Higher update frequency
   • Lower battery capacity
   • Larger energy harvesting rate or frame length
2. Adaptive Degree Distribution: For IDENTIFY, adaptive degree distribution provides limited improvement over fixed distribution
3. Energy Harvesting Effectiveness: Despite energy constraints, IRSA can achieve performance approaching ideal systems

Related Work

Random Access Protocols

• Traditional research focuses on minimizing PLR or maximizing throughput
• Recent attention to information freshness (AoI) metrics
• IRSA improves performance through time diversity and SIC decoding

Energy Harvesting Systems

• Previous research assumes receiver knows locations of discarded copies
• This paper first proposes a practically feasible identification method
• Validates feasibility of critical assumptions

Age of Information Research

• AoI serves as important metric for time-critical applications
• AoI analysis of random access protocols gradually gaining attention
• This paper extends to energy harvesting scenarios

Conclusions and Discussion

Main Conclusions

1. Proposes a practical method for receiver to identify transmission failures
2. IRSA under energy harvesting constraints can maintain good information freshness
3. Optimized degree distribution significantly improves AoI performance
4. IDENTIFY scheme outperforms conservative AVOID scheme

Limitations

1. Complexity: IDENTIFY scheme's decoding complexity is high, requiring testing all subsets of candidate lists
2. Assumptions: Relies on receiver's ability to accurately distinguish idle, singleton, and collision slots
3. Optimization Method: Uses heuristic Nelder-Mead algorithm, potentially converging to non-global optimum
4. Energy Model: Simplified energy harvesting model may differ from practical scenarios

Future Directions

1. Feedback Mechanisms: Explore use of receiver feedback
2. Complexity Reduction: Develop low-complexity copy identification algorithms
3. Asynchronous Scenarios: Extend to asynchronous IRSA
4. Practical Deployment: Validate in real IoT environments

In-Depth Evaluation

Strengths

Technical Innovation

1. Practical Breakthrough: First proposes practically feasible transmission failure identification method, validating important theoretical assumptions
2. Theoretical Contribution: Provides theoretical lower bound analysis of PLR under energy harvesting constraints
3. System Optimization: Comprehensive degree distribution optimization framework

Experimental Sufficiency

1. Comprehensive Comparison: Compares with multiple baseline methods
2. Parameter Sensitivity: Thorough analysis of parameter impacts on performance
3. Theoretical Validation: Simulation results verify theoretical analysis

Writing Quality

1. Clear Structure: Rigorous logic with clear hierarchy from problem definition to solutions
2. Mathematical Rigor: Rigorous theoretical analysis with complete proofs
3. Practical Value: Addresses critical issues in practical IoT deployment

Weaknesses

Method Limitations

1. Computational Complexity: Exponential complexity of IDENTIFY scheme limits practical application
2. Idealized Assumptions: Perfect slot synchronization and collision detection difficult to achieve in practice
3. Simplified Energy Model: Does not account for time-varying and correlated energy harvesting

Experimental Shortcomings

1. Scale Limitations: Only considers medium-scale systems (1000 devices)
2. Parameter Range: Relatively limited range for certain critical parameters
3. Lack of Practical Validation: Missing verification on real hardware platforms

Analysis Depth

1. Convergence Analysis: Insufficient convergence guarantees for optimization algorithm
2. Robustness: Limited analysis of robustness to parameter estimation errors

Impact

Academic Contribution

1. Theoretical Value: Provides new analytical framework for energy harvesting random access systems
2. Method Innovation: IDENTIFY algorithm extensible to similar problems
3. Practicality: Directly applicable to IoT system design

Application Prospects

1. IoT: Satellite communications, sensor networks and other large-scale deployment scenarios
2. Edge Computing: Status updates for resource-constrained devices
3. Industry 4.0: Real-time monitoring systems in manufacturing

Applicable Scenarios

Most Suitable Scenarios

1. Large-scale IoT: Large number of devices with limited energy
2. Time-critical Applications: Requiring maintained information freshness
3. Infrastructure-free Environments: Unable to frequently maintain or charge devices

Unsuitable Scenarios

1. Real-time Systems: Applications with extremely strict delay requirements
2. Small-scale Networks: Relatively large complexity overhead
3. Energy-sufficient Environments: Traditional methods already sufficient

References

The paper cites important works in related fields, including:

• Foundational IRSA protocol theory (Liva 2011)
• Theoretical framework for Age of Information (Yates et al. 2021)
• Previous research on energy harvesting systems (Demirhan & Duman 2019)
• Modern developments in random access protocols (Berioli et al. 2016)

This paper achieves good balance between theoretical analysis and practicality, providing valuable insights and solutions for protocol design in energy harvesting IoT systems. Despite some limitations, its core contributions have important implications for advancing the field.