Source-linked AI summary
Adaptable Semantic Compression and Resource Allocation for Task-Oriented Communications
Chuanhong Liu, Caili Guo, Yang Yang, Nan Jiang
TL;DR
The paper addresses semantic compression and limited-resource allocation in multi-user task-oriented communications. It proposes adaptable semantic compression with iterative compression-ratio/resource optimization, and reports improved task success alongside a higher-complexity user-selection extension.
Problem
Semantic compression reduces transmitted data, but limited wireless resources and users’ differing task priorities require coordinated resource allocation to preserve reliable task transmission.
Method
The paper develops an end-to-end task-oriented semantic communication architecture with adaptable semantic compression, then iteratively optimizes compression ratios and resource allocation under constraints.
Results
The proposed algorithms achieve higher average task success probability than baseline resource-allocation methods, while 80% semantic compression causes only tiny classification loss above 0 dB SNR.
Takeaways & Limitations
Optimal compression ratios depend on available bandwidth and power, so compression and resource allocation should be optimized jointly rather than selected independently.
Takeaways & Limitations
CRRAUS performs user selection through branch and bound, resulting in higher algorithm complexity than CRRA.
Abstract
from arXiv · showhide
Task-oriented communication is a new paradigm that aims at providing efficient connectivity for accomplishing intelligent tasks rather than the reception of every transmitted bit. In this paper, a deep learning-based task-oriented communication architecture is proposed where the user extracts, compresses and transmits semantics in an end-to-end (E2E) manner. Furthermore, an approach is proposed to compress the semantics according to their importance relevant to the task, namely, adaptable semantic compression (ASC). Assuming a delay-intolerant system, supporting multiple users indicates a problem that executing with the higher compression ratio requires fewer channel resources but leads to the distortion of semantics, while executing with the lower compression ratio requires more channel resources and thus may lead to a transmission failure due to delay constraint. To solve the problem, both compression ratio and resource allocation are optimized for the task-oriented communication system to maximize the success probability of tasks. Specifically, due to the nonconvexity of the problem, we propose a compression ratio and resource allocation (CRRA) algorithm by separating the problem into two subproblems and solving iteratively to obtain the convergent solution. Furthermore, considering the scenarios where users have various service levels, a compression ratio, resource allocation, and user selection (CRRAUS) algorithm is proposed to deal with the problem. In CRRAUS, users are adaptively selected to complete the corresponding intelligent tasks based on branch and bound method at the expense of higher algorithm complexity compared with CRRA. Simulation results show that the proposed CRRA and CRRAUS algorithms can obtain at least 15% and 10% success gains over baseline algorithms, respectively.
I. INTRODUCTION
Task-oriented communication transmits task-relevant meaning rather than every bit, but semantic compression and resource allocation remain challenging because semantics depend on the task and wireless resources are limited. This paper develops an E2E multi-user architecture with adaptable semantic compression and optimization methods for compression ratios and resource allocation.
- Motivation: Task-oriented communication transmits meaningful raw-data content for intelligent tasks instead of requiring accurate reception of every bit.The paradigm is intended to reduce network traffic and alleviate spectrum shortage.
- Open challenges: Semantic compression is task-dependent because semantic information is abstract and subjective, creating a challenge for unified compression criteria.The same data may have different semantics across intelligent tasks.
- Open challenges: Semantic compression reduces transmitted data and computing overhead, while resource allocation is needed to address limited wireless resources and prioritized reliability requirements.The paper identifies resource management as necessary because compression alone does not resolve wireless-resource limitations.
- Proposed framework: The proposed E2E framework extracts, compresses, and transmits raw-data semantics to an edge server, which performs intelligent computing and returns task results.The system uses semantic and channel encoders at the transmitter and corresponding decoders at the receiver.
- Adaptable semantic compression: Adaptable semantic compression transmits semantic features whose importance exceeds a compression-ratio-determined threshold and discards the others.Existing semantic communication systems are represented as the special case ω0 = 0.
- Optimization objective: The system improves success transmission probability by jointly optimizing semantic compression ratios and wireless resource allocation.Success transmission probability is mainly affected by power, bandwidth, and semantic compression ratio.
B. Problem Formulation
The paper defines task-level success probability to evaluate semantic communication and formulates joint optimization of resource allocation, compression ratios, and user selection under wireless and compression constraints.
- The success probability of tasks evaluates semantic communication by jointly reflecting transmission and semantic understanding.It controls the tradeoff between semantic transmission and semantic understanding.
- Users may receive different service levels, with higher-priority users requiring performance protection.For example, fire-detection users in smart factories may require highly reliable, low-latency communication.
- Only a subset of users may be selected because limited wireless resources must preserve higher-priority user performance.The selection index β_i indicates whether user i is selected.
- The optimization jointly chooses resource allocation, compression ratios, and user selection to maximize weighted sum success probability under system constraints.Constraints cover total and minimum bandwidth, total and minimum transmit power, compression ratios, and service-level assignment.
III. E2E SEMANTIC COMMUNICATION AND ASC
This section presents the end-to-end semantic communication architecture, its loss function, semantic-feature importance evaluation, and the relationship between task performance and compression ratio.
- The section develops an end-to-end semantic communication system and uses semantic-feature importance as the basis for adaptable semantic compression.It also studies how intelligent-task performance varies with compression ratio.
A. E2E Semantic Communication Network Design
The proposed network uses jointly trained neural semantic transceivers and a loss that balances task performance with transmitted-semantic mutual information.
- Deep neural networks implement a joint encoder-decoder framework for capturing and transmitting meaningful source semantics.The framework targets complex source data, including image recognition tasks.
- The semantic transceiver is jointly trained end-to-end with a neural-network channel model under an AWGN channel assumption.Gradients are backpropagated from the semantic decoder output to the semantic encoder input.
- B. Loss Function Design: The loss combines task loss with a weighted mutual-information upper-bound term to improve task performance while increasing achieved data rate.The task term can use cross-entropy or triplet loss, while κ weights mutual information between transmitted and received semantics.
- B. Loss Function Design: The mutual-information upper bound is optimized using a variational distribution qθ(y|x) implemented by neural networks.The bound exceeds the true mutual information, and its variational objective is minimized by gradient descent.
- B. Loss Function Design: Training uses two phases: unsupervised estimation of the achieved data rate, followed by joint-system training with the combined loss.Both phases use mini-batch gradient descent, and adaptable semantic compression is applied only during inference.
C. Semantic Importance Evaluation
The paper evaluates semantic-feature importance according to task relevance and uses gradient-based weights that can be shared between sender and receiver for adaptable compression.
- Semantic importance is defined as the correlation between semantic features and the intelligent task.Features irrelevant to the task represent semantic redundancy that can be further compressed.
- A gradient-based method measures feature importance by differentiating a semantic concept activation with respect to feature activations.The gradients are globally averaged over the feature-map width and height to obtain importance weights.
- The resulting weight captures the semantic importance of a feature for a semantic concept through a partial linearization of the downstream network.It is computed from successive products of network weights and activation-function gradients.
- Importance weights depend only on network parameters, so sender and receiver can store them as shared task-specific knowledge without transmitting feature-map indices.The paper uses the gradient-based method to calculate these weights.
D. Intelligent Task Performance Model
The paper models task performance as a function of semantic compression ratio because neural-network behavior makes a closed-form relationship intractable. It estimates this relationship from feature-map importance and fits an exponential function before optimizing compression and resources.
- A closed-form task-performance function η(o) is intractable because it depends on neural-network behavior.
- The fitting parameters ζ are learned iteratively from sampled compression-ratio and task-performance points.
- Feature-map importance is used to remove unimportant maps progressively and measure corresponding task performance and compression ratios.
- The estimated mapping is fitted with η(o) = ζ1e^ζ2o + ζ3e^ζ4o.
- For equal-service-level users, all users are selected and the optimization focuses on compression ratios and resource allocation.
- Because the objective is non-concave, CRRA separates the problem into two subproblems and solves them iteratively until convergence.
A. Optimal Compression Ratios
With resource allocation fixed, each user’s compression ratio can be optimized independently by one-dimensional enumeration. The resulting ratios are then used in an SCA-based resource-allocation procedure, and the two stages are iterated to convergence.
- With a fixed resource-allocation policy, each user’s optimal semantic compression ratio is independent of the others.
- Because o_i lies between 0 and 1, one-dimensional enumeration searches candidate ratios to obtain each user’s optimum.
- The resource-allocation stage treats η(o_i) as a constant α_i once compression ratios are fixed.
- Slack variables and successive convex approximation transform the resource-allocation formulation into a convex problem.
- The convex resource-allocation problem is solved iteratively, updating variables until the objective converges.
- CRRA alternates compression-ratio enumeration with resource allocation and repeats until the objective value converges.
V. CRRAUS ALGORITHM
CRRAUS addresses users with different priorities by adapting user selection to wireless resources and service levels. It extends the optimization framework to scenarios where only part of the users can be selected.
- CRRAUS targets users with different task priorities and consequently different service levels.
- Only part of the users may be selected because wireless resources are constrained.
- The algorithm adaptively adjusts user selection according to wireless resources and service levels.
A. Algorithm Design
The algorithm design decomposes the nonconvex problem into compression-ratio, user-selection, and resource-allocation subproblems. These are solved with enumeration, branch and bound, and SCA-based optimization in an iterative procedure.
- The original problem is difficult because its objective function is non-concave and its constraints are nonconvex.
- CRRAUS first optimizes compression ratios with fixed resources and selection, then solves user selection, and finally optimizes resource allocation.
- Compression-ratio optimization maximizes each user’s weighted task-success probability and uses one-dimensional enumeration.
- User selection is formulated as a 0-1 integer program and solved with branch and bound.
- Branch and bound branches on fractional variables and updates upper and lower bounds until an integer solution is obtained.
- With selection and compression ratios fixed, resource allocation uses SCA and the dual method, while all three subproblems are iterated to convergence.
B. Convergence and Complexity Analysis
CRRA is shown to converge to a KKT-satisfying point while combining compression-ratio optimization with resource allocation at stated computational cost. CRRAUS extends the framework with user selection through branch and bound, increasing algorithmic complexity.
- Convergence: CRRA’s total success probability is monotonically non-decreasing, and its iterates converge to a point satisfying the original problem’s KKT conditions.The convergence analysis focuses on the resource-allocation subproblem, while compression ratios are solved by one-dimensional enumeration.
- CRRA complexity: CRRA has total complexity O(T0KU + T0U^3.5), combining compression-ratio enumeration with resource-allocation optimization across T0 iterations.With fixed resource allocation, compression-ratio optimization costs O(KU); with fixed compression ratios, resource allocation costs O(U^3.5).
- CRRAUS complexity: CRRAUS adds one-dimensional enumeration, branch-and-bound user selection, and SCA-based optimization to its computational complexity.The algorithm addresses compression ratios, user selection, and resource allocation through separate optimization components.
- Simulation results: At 80% semantic compression, classification-accuracy loss is tiny above 0 dB SNR while transmitted data and delay are reduced.The comparison uses VGG and ResNet backbones and contrasts ASC with traditional communication and lower compression levels.
- Simulation results: Optimal compression ratios vary with available bandwidth and transmit power, reflecting a trade-off between transmission efficiency and task performance.The reported optimum is 0.7 at 10 MHz and 0.8 at 20 MHz bandwidth.
- Simulation results: CRRA outperforms FCR, FRA, and MSR in task success probability, especially with many users, while incurring additional computational complexity.The advantage is attributed to jointly choosing compression ratios and resource allocation under delay constraints.
VII. CONCLUSION
The paper develops task-oriented multi-user semantic communication with adaptive semantic compression, then jointly optimizes compression ratios and resource allocation to improve task success.
- The proposed architecture adaptively compresses semantics to reduce transmitted data and communication burden.
- CRRA iteratively optimizes resource allocation and compression ratios for users with the same service levels.
- CRRAUS adaptively selects users with varying service levels using a branch and bound method.
- Both CRRA and CRRAUS achieve higher task success probability than benchmark algorithms in simulations.
APPENDIX A Proof of Lemma 1
The appendix proof derives the stated lemma from preceding equations and concludes the argument using the channel-coefficient distribution.
- The proof begins by using equations (6) and (7) to establish the lemma's intermediate result.
- The final equality follows from the assumption hi ∼N(0, δ2).
- The appendix closes by declaring the proof of Lemma 1 complete.