580 California St., Suite 400
San Francisco, CA, 94104
Key research themes
1. How does the performance and optimization of the JVM interpreter affect embedded Java execution efficiency?
This research area concentrates on understanding and improving the execution efficiency of Java bytecode on embedded systems by analyzing the performance bottlenecks within the Java Virtual Machine (JVM) interpreters. It is critical because embedded Java environments often have limited hardware resources and require optimal runtime performance for real-time and embedded applications. By exploring the effects of different JVM builds, compiler toolchains, and CPU interactions, researchers aim to reduce overheads and enhance interpreter performance, ultimately improving responsiveness and energy efficiency in embedded contexts.
Key finding: This study evaluated the impact of different GNU GCC compiler versions on JVM interpreter performance across multiple OpenJDK releases using the Renaissance Benchmark Suite. It found statistically significant performance... Read more
Key finding: By employing a multifaceted profiling methodology that combines high-level object metrics with low-level microarchitectural counters within MaxineVM, this paper characterized memory behaviors of Java benchmarks... Read more
Key finding: This research demonstrated that reorganizing Java objects during garbage collection using Breadth First (BF), Depth First (DF), and optimized ordering schemes can yield 3-16% improvements in execution time and reduce cache... Read more
2. What are the design and implementation paradigms enabling effective embedded Java development for real-time and resource-constrained systems?
This theme explores specialized software engineering methodologies, architectural frameworks, and programming language extensions that support embedded Java development in real-time and resource-limited environments. The objective is to understand how object-oriented paradigms, component-oriented designs, and real-time Java extensions can be harnessed for embedded system constraints such as timely execution, memory footprint, and system safety. These paradigms enable maintainable, reusable, and modular embedded Java applications suitable for critical systems including manufacturing and robotics.
Key finding: The paper advocates embedding software engineering principles and abstraction methods tailored to embedded systems characterized by constraints on memory, power, and timing. It underscores the growing predominance of software... Read more
Key finding: This study presents a Java PLC architecture integrating Petri Nets and Sequential Function Charts with real-time Java, enabling concurrent execution and control of manufacturing processes. It introduces a component-oriented... Read more
Key finding: The paper introduces DECCA, a Java API and methodology for developing synchronized, concurrent component-based systems with formal verification capabilities. It enables high-level behavioral modeling using Java’s OOP features... Read more
3. How do language-level features and interoperability influence the adaptability and security of embedded Java systems?
This theme investigates the adaptation of new Java language constructs (e.g., generics, lambda expressions), cross-language interoperability mechanisms, and security enforcement models to embedded Java environments. It includes the practical implications of integrating functional programming paradigms, enabling runtime interaction with heterogeneous hardware, and protecting mobile code and bytecode from attacks, which are essential concerns in embedded systems where resource constraints and security risks coexist.
Key finding: This paper proposes an approach to Java generics inspired by C++ templates to reduce runtime overhead by generating separate typed Java source code through instantiation, improving performance. The technique mitigates type... Read more
Key finding: Through empirical analysis of the top 50 Java APIs, this study finds that only a minority have adopted functional interfaces and lambda support, revealing a cautious ecosystem transition. The paper discusses barriers and... Read more
Key finding: This work analyzes challenges in achieving cross-language interoperability of heterogeneous hardware-accelerated code generated via TornadoVM on the JVM, proposing modifications to support 'code-interoperability-mode.' The... Read more
Key finding: This paper presents an access control framework integrating the Ponder policy language with Java security to enforce flexible, dynamic permissions on mobile Java code. The framework enables separation of policy specification... Read more
Key finding: The authors propose a system employing a thin hypervisor-based trust zone to execute encrypted Java bytecode securely, addressing the key management challenges posed by JVM interpretation. By integrating a protected... Read more