Uplink Scheduling via Convex Optimization
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing scheduling methods in cellular radio systems, particularly WCDMA, face challenges such as non-optimal throughput, instability, and coverage loss due to ad hoc design and the difficulty in computing solutions in real time, leading to suboptimal performance and inefficient resource allocation.
Innovation Solution
The method involves forming a cost function based on the load factor for each user equipment, approximated by a quadratic function, to maximize throughput through convex optimization of the sum of individual cost functions, and scheduling grants accordingly, while considering minimum and maximum load allocation, ranking, and adaptive weighting to ensure fairness and stability.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If ad hoc scheduling methods are used to allocate uplink resources, then implementation simplicity is maintained, but throughput is suboptimal and resource allocation is inefficient
Solution Approach 1:
The patent transforms the scheduling problem from grant-based allocation to load factor-based allocation. By changing the scheduling parameter from discrete grant levels to continuous load factor values, the system achieves optimal throughput through convex optimization while maintaining computational efficiency through quadratic cost functions.
Solution Approach 2:
The patent replaces the traditional mechanical scheduling approach (step-by-step grant adjustment) with a mathematical optimization framework. The convex optimization algorithm with quadratic cost functions substitutes the ad hoc scheduling mechanism, providing systematic and optimal resource allocation.
2Productivity
If conventional scheduling methods are used, then computational simplicity is maintained, but real-time computation capability is insufficient
Solution Approach 1:
By formulating the scheduling problem with quadratic cost functions and convex constraints, the patent enables real-time computation through efficient convex optimization algorithms. The parameter transformation from grant-based to load factor-based scheduling allows the use of numerically stable and computationally efficient solution methods.
3Productivity
If maximum traffic is scheduled to increase capacity, then cell capacity is improved, but coverage and stability are compromised due to excessive interference
Solution Approach 1:
The patent implements a closed-loop scheduling system where the load factor serves as a feedback parameter. The convex optimization algorithm continuously adjusts uplink grants based on the current load factor, ensuring that capacity is maximized while maintaining stability. The quadratic cost functions inherently penalize excessive load allocation, providing automatic feedback control.
Solution Approach 2:
By using load factor as the scheduling parameter instead of discrete grants, the system achieves continuous and fine-grained control over uplink resource allocation. This parameter transformation enables precise control of interference levels while maximizing cell capacity through convex optimization.
4Productivity
If discrete grant levels are used for scheduling control, then standard compliance is maintained, but resource allocation precision is limited
Solution Approach 1:
The patent transforms the discrete grant parameter into a continuous load factor parameter. This parameter change enables precise and fine-grained resource allocation through convex optimization, overcoming the granularity limitations of discrete grant levels while maintaining standard compliance through proper mapping back to grant values.
Data Source
AI summary
In methods and devices for scheduling uplink transmission in a cellular radio system for a number of user equipments transmitting data over an air-interface each user equipment is associated with an individual uplink load factor. Further a cost function is formed based on the load factor for each user equipment that is to be scheduled for uplink transmission, wherein the cost function is approximated by a quadratic function. The throughput of all scheduled user equipments is maximized using a convex optimization of the sum of the individual cost functions, and the grant for uplink transmission is scheduled in accordance with the optimized cost functions.