Deep Multi-Armed Bandit for OFDM-IM Subcarrier Selection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing OFDM-IM systems face challenges in optimizing average bit error probability due to the exponential increase in complexity with the number of subcarriers, leading to reduced energy efficiency and suboptimal bit error rates, as they divide subcarriers into groups, affecting channel size and combinatorial effects.
Innovation Solution
A deep multi-armed bandit framework is introduced to jointly select subcarriers, using a neural network to minimize bit error probability by representing subcarrier selection patterns as vectors and employing the upper confidence bound algorithm to maximize reward, without storing combinatorial observations, thus overcoming memory limitations and individual subcarrier selection limitations.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If maximum likelihood detection is used to detect subcarrier indices, then detection accuracy is improved, but computational complexity increases exponentially with the number of subcarriers
Solution Approach 1:
The patent divides the set of N subcarriers into G groups, where each group contains N/G subcarriers. This segmentation reduces the detection complexity from exponential in N to exponential in N/G, making the system feasible for large N. The receiver performs maximum likelihood detection separately within each group rather than over all subcarriers simultaneously.
Solution Approach 2:
The patent introduces a group dimension by organizing subcarriers into G groups. The detection problem is transformed from selecting K subcarriers out of N to selecting K subcarriers from G groups, where each group contributes at most one subcarrier. This dimensional restructuring reduces the search space from C(N,K) to C(G, K/G) approximately.
2Device complexity
If subcarriers are divided into groups to reduce complexity, then computational complexity is reduced, but energy efficiency decreases due to increased number of activated subcarriers
Solution Approach 1:
The patent changes the detection parameter from individual subcarrier selection to group-based selection. By organizing subcarriers into groups and selecting one subcarrier per group, the system maintains lower complexity while reducing the total number of activated subcarriers compared to individual selection, thereby improving energy efficiency.
Solution Approach 2:
The patent applies different selection criteria to different groups of subcarriers. Each group is evaluated based on local channel conditions, and the best subcarrier within each group is selected. This local optimization approach ensures that only necessary subcarriers are activated, improving energy efficiency while maintaining detection performance.
3Ease of operation
If individual subcarrier selection is performed based on channel information, then selection simplicity is improved, but average bit error probability increases due to ignoring combinatorial effects
Solution Approach 1:
The patent employs feedback mechanisms where the receiver sends back information about detected subcarrier indices and channel conditions. The transmitter uses this feedback to adjust future subcarrier selections, taking into account the combinatorial effects of selected subcarriers. This feedback loop enables the system to optimize bit error probability while maintaining selection simplicity.
Solution Approach 2:
The patent performs preliminary evaluation of subcarrier groups based on channel state information before actual transmission. By pre-assessing which groups contain the best subcarriers and how they combine, the system can make informed selection decisions that account for combinatorial effects, improving reliability without sacrificing operational simplicity.
Data Source
AI summary
A method and apparatus for optimizing average bit error probability via a deep multi-armed bandit in an orthogonal-frequency division multiplexing and index modulation system for low power communication are proposed. The method proposed in the present invention comprises: detecting BPSK symbols and subcarriers among all subcarriers; defining a combination of selected subcarriers as a subcarrier selection pattern; selecting the subcarrier selection pattern through learning to minimize the average bit error probability for all combinations of selected subcarriers; and updating a learning parameter of the subcarrier selection pattern selected through learning.