Defense Notices

Genetic Algorithms (GAs), a subclass of evolutionary algorithms, seek to apply the concept of natural selection to promote the optimization and furtherance of “something” designated by the user. GAs generate a population of chromosomes represented as value strings, score each chromosome with a “fitness function” on a defined set of criteria, and mutate future generations depending on the scores ascribed to each chromosome. In this project, each chromosome is a bitstring representing one canvased color artwork. Artworks are scored with a variety of design fundamentals and user preference. The artworks are then evolved through thousands of generations and the final piece is computationally drawn for analysis. While the rise of gradient-based optimization has resulted in more limited use-cases of GAs, genetic algorithms still have applications in various settings such as hyperparameter tuning, mathematical optimization, reinforcement learning, and black box scenarios. Neural networks are favored presently in image generation due to their pattern recognition and ability to produce new content; however, in cases where a user is seeking to implement their own vision through careful algorithmic refinement, genetic algorithms still find a place in visual computing.

Wireless edge systems typically need to complete timely computation and inference tasks under strict power, bandwidth, latency, and processing constraints. As AI models and datasets grow in size and complexity, the traditional model of sending all data to a remote cloud or running full inference on edge device becomes impractical. This creates a need for communication-computation co-design to enable efficient AI task processing at the wireless edge. To address this problem, we investigate task-aware communication-computation optimization for two specific problem settings.

First, we explore semantic communication that transmits only the information essential for the receiver’s computation tasks. We propose a semantic-aware and goal-oriented communication method for object detection. Our proposed approach is built upon the auto-encoders, with the encoder and the decoder are respectively implemented at the transmitter and receiver to extract semantic information for the specific computation goal (e.g., object detection). Numerical results show that transmitting only the necessary semantic features significantly improves the overall system efficiency.

Second, we study collaborative inference in wireless edge networks, where energy-constrained devices aim to complete delay-sensitive inference tasks. The inference computation is split between the device and an edge server, thereby achieving collaborative inference. We formulate a utility maximization problem under energy and delay constraints and propose Bayes-Split-Edge, which uses Bayesian optimization to determine the optimal transmission power and neural network split point. The proposed framework introduces a hybrid acquisition function that balances inference task utility, sample efficiency, and constraint violation penalties. We evaluate our approach using the VGG19 model, the ImageNet-Mini dataset, and real-world mMobile wireless channel datasets.

Overall, this research is aimed at developing efficient edge AI systems by incorporating the underlying wireless communications limitations and challenges into AI tasks processing.