LoRa MAC Compression Using Sparse Vectors to Cut Packet Collisions
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
LoRa-based LPWAN technologies face limitations in throughput due to collisions during simultaneous transmissions and stringent duty-cycle restrictions, leading to low data rates and network congestion, which are not adequately addressed by existing MAC protocols.
Innovation Solution
Integration of compressive sensing techniques within the LoRa MAC layer to compress data into sparse vectors, allowing for efficient transmission and recovery of sensor signals, reducing the need for retransmissions and optimizing data exchange by combining current measurements with past data in a linear combination, thereby enhancing data rate and reliability.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of operation
If LoRa devices use pure Aloha MAC scheme for simultaneous transmissions, then devices can communicate at any time without strict timing constraints, but network congestion and collisions increase inevitably
Solution Approach 1:
The patent implements dynamic slot allocation where time slots are not fixed but adaptively assigned based on traffic conditions and device priorities. The MAC layer dynamically adjusts the number and duration of slots, allowing the system to transition between contention-based and scheduled access modes to optimize both flexibility and throughput under different network conditions.
Solution Approach 2:
The patent segments the transmission medium into multiple time slots and priority levels, dividing the network traffic into different classes that can be handled separately. This segmentation allows simultaneous transmissions at different priority levels without complete collisions, improving overall network productivity while maintaining operational flexibility for each segment.
2Reliability
If LoRa devices transmit data frequently to ensure data delivery, then data reliability improves, but energy consumption increases and battery life decreases
Solution Approach 1:
The patent implements preliminary actions through proactive data buffering and predictive transmission scheduling. Devices prepare data in buffers before transmission is needed and the MAC layer schedules transmissions in advance based on predicted traffic patterns and channel conditions, reducing the need for frequent retransmissions and energy-consuming emergency transmissions.
Solution Approach 2:
The patent implements feedback mechanisms where devices and the network server exchange acknowledgment signals and channel quality information. This feedback allows the system to adjust transmission strategies, retry only failed packets, and optimize transmission timing, improving packet delivery ratio while minimizing unnecessary retransmissions and energy consumption.
3Use of energy by moving object
If LoRa devices use Class A communication mode for energy efficiency, then battery life extends, but data exchange rate and network responsiveness are limited
Solution Approach 1:
The patent creates a universal MAC layer framework that can operate in multiple modes including Class A, Class B, and Class C behaviors. The system can dynamically switch between energy-efficient modes and high-throughput modes depending on application requirements, making the same hardware capable of serving both battery-constrained and performance-critical applications without dedicated hardware for each mode.
Solution Approach 2:
The patent implements dynamic mode switching within Class A operations, allowing devices to transition between sleep mode and active transmission mode based on real-time network conditions and data priority. High-priority data can trigger immediate wake-up and transmission, while low-priority data waits for scheduled slots, dynamically balancing energy efficiency and data exchange rate without requiring separate hardware configurations.
4Area of stationary object
If the number of LoRa devices in the network increases to expand coverage, then network area expands, but collision probability increases and individual device throughput decreases
Solution Approach 1:
The patent segments the network into multiple virtual channels and priority queues, allowing devices to be distributed across different logical segments rather than competing on a single shared medium. This segmentation reduces collision probability even as the number of devices increases, maintaining per-device throughput while expanding overall network capacity and coverage area.
Solution Approach 2:
The patent introduces additional dimensions for resource allocation beyond simple time slots, including frequency sub-channels, priority levels, and spatial grouping. This multi-dimensional resource allocation allows the network to accommodate more devices by distributing them across multiple dimensions, reducing collisions in each dimension while maintaining high per-device throughput and expanding overall network capacity.
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
The proposed solution significantly improves packet delivery ratio and throughput by reducing the number of transmitted packets while maintaining low complexity, enabling more efficient data exchange and longer battery life in LoRa devices.
Implementation Method 1
compress the chirp modulated encoded bit stream of the first transmission packet into sparse vectors
Implementation Method 2
modulate the encoded bit stream by chirp spread spectrum modulation to generate a chirp modulated encoded bit stream
Data Source
AI summary
A LoRa device for communicating sensor signals in a low power wide area network (LPWAN) includes a physical layer using Hamming encoding and Gray indexing with chirp spread signal (CSS) modulation to encode and modulate the sensor signals and a medium access layer (MAC) including a compressive sensing sub-layer which reduces encoded, modulated signals to sparse vectors. A transmission packet is formed by combining the sparse vectors with a selected set of sparse vectors representing past measurements and the incoming velocity of the sensor signals. A receiver decompresses the transmission packet by reconstructing, at a sparse recovery sub-layer of a receiver MAC layer, the encoded, modulated sensor signals. A decoder path removes the CSS modulation and Gray indexing, and Hamming decodes the sensors signals.