Dynamic Resource Allocation in Heterogeneous Wireless Backhaul
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
In the Evolved Packet System (EPS) with LTE networks, managing backhaul link resources on a per-link basis leads to congestion in some links and underutilization in others, resulting in inefficient network asset usage.
Innovation Solution
A point-multipoint (PMP) node identifies traffic conditions and achievable throughput for each link, assigns priorities based on these metrics, and allocates resources to the highest-priority links to optimize traffic transmission.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of operation
If resources are managed on a per-link basis, then each link can be independently controlled, but some links become over-utilized while others are under-utilized
Solution Approach 1:
The patent merges multiple backhaul links into link aggregates, treating them as a single logical resource pool. The PMP node dynamically allocates traffic across member links based on real-time conditions, combining previously independent links into a unified resource management system that prevents both over-utilization and under-utilization of individual links.
Solution Approach 2:
The system implements dynamic resource allocation where the PMP node continuously monitors link conditions and adjusts traffic distribution in real-time. This dynamic approach allows the system to adapt to changing network conditions, shifting traffic from congested links to available capacity on other links within the aggregate, thereby optimizing overall network asset utilization.
2Stability of the object's composition
If resources are allocated statically to each link, then resource allocation is simple and stable, but congestion occurs in high-traffic links and waste in low-traffic links
Solution Approach 1:
The patent transitions from static per-link resource allocation to dynamic aggregate-based allocation. The PMP node continuously assesses link conditions and reallocates traffic dynamically, maintaining stability through controlled adjustments rather than rigid fixed assignments. This dynamic approach ensures high-traffic links receive adequate capacity while preventing waste on low-utilization links.
Solution Approach 2:
The system changes the allocation parameters from fixed per-link bandwidth assignments to flexible traffic distribution based on real-time link conditions. The PMP node modifies traffic engineering parameters dynamically, adjusting the amount of traffic routed through each member link based on current utilization, latency, and other performance metrics.
3Productivity
If more links are added to increase capacity, then network throughput increases, but resource management complexity increases and inefficiency persists
Solution Approach 1:
The patent combines multiple backhaul links into logical aggregates, reducing management complexity by treating multiple physical links as a single resource pool. The PMP node performs centralized resource allocation for the entire aggregate, simplifying control compared to managing each link independently, while still utilizing the combined capacity of all member links to increase overall throughput.
Solution Approach 2:
The link aggregate structure creates a universal resource pool that can serve multiple base stations and handle various types of traffic. The PMP node implements a unified resource allocation mechanism that works across all member links regardless of their individual characteristics, providing multi-functional resource management that handles both capacity allocation and load balancing.
Data Source
AI summary
A system to receive an instruction to allocate resources to a particular link of a group of links, to permit traffic to be transmitted to one of a group of nodes, associated with one of a group of base stations, via the particular link; and identify a respective achievable throughput, associated with each link during a first time period, based on a respective capacity and condition associated with each link. The system is also to identify a respective achieved throughput, associated with each link, based on a respective achieved throughput associated with each link during a second time period and a respective quantity of traffic to be transmitted via each link during the first time period; generate a respective value, associated with each link, based on the respective achievable throughput and the respective achieved throughput; and transmit, to a node and via a link, a quantity of traffic, based on a determination that a respective value, associated with the link, is a greatest respective value.