.png){kind=icon}
Return
A common software framework for multi-function radar via three-level dependency decoupling
.png){kind=icon}
PRE
AI
DOI：10.1016/j.sysarc.2026.104021.png)
.png){kind=icon}
Original request for help.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
Modern multi-function radar (MFR) systems execute heterogeneous missions across diverse platforms, demanding a common software framework that enables systematic reuse across the product family. However, prevailing defense software practices remain anchored to monolithic architectures in which reusable core assets are inseparably entangled with volatile, project-specific modules. Transforming such codebases into a reusable framework has remained elusive due to three mutually reinforcing challenges: the absence of a quantitative basis to govern inter-module dependencies at the macro level, compile-time static I/O binding at the boundary level, and mutex-based synchronization at the micro level. We propose a framework that resolves inter-module coupling across three levels through three synergistic strategies. First, a volatility-dependency-based layered architecture enforces a unidirectional dependency flow that isolates invariant core assets from volatile modifications. Second, a protocol-agnostic dynamic I/O interface driven by an inversion-of-control mechanism achieves zero-code adaptability across heterogeneous environments. Third, RiPPS, a lightweight intra-process publish–subscribe middleware, guarantees anonymous communication and deadlock-free concurrency via a memory-isolated stream. Evaluation on two operational MFR subsystems demonstrates reductions of up to 51% in modified cyclomatic complexity, 42% in coupling references, and 56% in lines of code, while concurrency benchmarks confirm that RiPPS preserves publish–subscribe anonymity at latencies competitive with bare lock-free queues.
Keywords:
Multi-function radar
Software product line
Common software framework
Layered architecture
Intra-process communication
Journal
IF:
4.1
Papers:
3.0K
Citations:
4.2K
Organization
Cited Papers
No cited papers available