Distributed MIMO Power Allocation for Equal Throughput

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

In distributed MIMO cellular radio networks, existing methods for allocating transmit power fail to achieve equal throughput for each mobile terminal while maximizing network throughput, often resulting in coverage penalties, especially when multiple spatial modes serve multiple mobile terminals.

Innovation Solution

The method involves selecting groups of mobile terminals, performing channel measurements, calculating maximum equal throughput, and deriving power allocations for each spatial mode or beam to ensure equal throughput across the network, thereby maximizing the number of supported terminals without coverage penalties.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Productivity

If water-pouring power allocation is used to maximize MIMO channel capacity, then capacity is significantly improved, but all transmitter power is allocated to one spatial mode causing coverage penalty

Engineering Contradiction:
ImproveMIMO channel capacityVSAvoidcoverage area
Core Design Contradiction:
ProductivityVSArea of stationary object

Solution Approach 1:

The patent applies local quality by allocating different power levels to different spatial modes based on their specific channel conditions. Instead of concentrating all power in one mode (water-pouring), the system tailors power distribution to local channel characteristics, ensuring each spatial mode receives appropriate power to maintain coverage while contributing to overall capacity.

Inventive Principle:
Principle #3Local quality

Solution Approach 2:

The patent changes the power allocation parameter from the water-pouring approach (all-or-nothing allocation) to a distributed allocation scheme where power is spread across multiple spatial modes. This parameter change enables simultaneous achievement of high capacity and broad coverage by adjusting how total transmitter power is distributed among available spatial modes.

Inventive Principle:
Principle #35Parameter changes

2Ease of operation

If equal throughput is achieved for all mobile terminals in distributed MIMO, then fairness is improved, but network throughput may be reduced compared to optimized allocation

Engineering Contradiction:
Improvethroughput fairnessVSAvoidnetwork throughput
Core Design Contradiction:
Ease of operationVSProductivity

Solution Approach 1:

The patent implements dynamic power allocation that adapts to channel conditions while maintaining equal throughput constraints. The system dynamically adjusts power distribution across spatial modes and time slots to ensure fairness, rather than using static equal allocation, thereby preserving network throughput while achieving throughput equality among terminals.

Inventive Principle:
Principle #15Dynamics

Solution Approach 2:

The system employs periodic scheduling and power reallocation to maintain equal throughput across terminals. By periodically adjusting power allocation based on accumulated throughput measurements and channel conditions, the system ensures fairness over time while maximizing overall network utilization through efficient resource scheduling.

Inventive Principle:
Principle #19Periodic action

Data Source

PatentUS7702353B2Transmit power allocation in a distributed MIMO system
Publication Date: 2010.04.20 APPLE INC
  • US7702353B2 patent drawing
  • US7702353B2 patent drawing
  • US7702353B2 patent drawing

AI summary

Allocate power so as to maximise the throughput of each user of a multi-user MIMO group, with the constraint that over time all users in the group have equal throughput. This differs from equal capacity per slot in that each user may be assigned multiple slots as well as unequal power. This is illustrated in FIG. 4. Total throughput is maximised on any given slot for any two users. Power is shared between the spatial modes such that the total number of slots used by the two users is minimised. The membership of the MIMO group may change between slots and thus throughput is not necessarily equalised on a slot by slot basis.