Deterministic Scheduling for WIA-PA Industrial Wireless Networks

The patent develops specialized scheduling algorithms tailored to the specific characteristics of WIA-PA networks, including the TDMA communication mechanism and dual-channel structure. The backtracking method and least slack first method are designed with parameters and logic specific to WIA-PA network topology and transmission requirements, ensuring local optimization for this network type rather than applying generic scheduling approaches.

The patent modifies scheduling parameters and algorithms to match WIA-PA network characteristics. The backtracking method adjusts search strategies based on network state, and the least slack first method calculates time margins specific to WIA-PA's TDMA framework. These parameter adaptations enable the scheduling system to effectively handle WIA-PA's unique communication patterns and resource constraints.

The patent segments the scheduling problem into two distinct approaches based on network scale: backtracking method for middle and small scale networks where exhaustive search is feasible, and least slack first method for large scale networks where approximation is necessary. This segmentation allows each algorithm to be optimized for its appropriate domain, balancing success rate and computational complexity according to network size.

The patent implements a dynamic scheduling strategy that adapts the algorithm selection based on network scale. The system dynamically chooses between optimal backtracking and suboptimal least slack first methods according to the number of nodes and communication requirements, allowing the computational complexity to scale appropriately with network size while maintaining acceptable success rates.

The least slack first method implements a partial search strategy that focuses computational effort on the most critical scheduling decisions (those with smallest time margins) rather than exhaustively evaluating all possibilities. This partial action approach achieves acceptable success rates for large scale networks while dramatically reducing computational time compared to optimal methods.

The patent calculates time margins for all data streams in advance and uses these preliminary calculations to prioritize scheduling decisions. By pre-computing urgency metrics and sorting data streams by least slack first, the system prepares scheduling information beforehand, enabling rapid decision-making during the actual scheduling process without exhaustive real-time computation.

The patent creates a dynamic framework that adjusts the degree of optimization based on network requirements. For time-critical applications, the system can use the faster least slack first approximation, while for applications requiring maximum reliability, the backtracking optimal method is employed. This dynamic approach allows the system to balance computational feasibility and solution optimality according to specific application needs.

The patent modifies the search depth and pruning parameters of the backtracking algorithm to control the trade-off between optimality and computational feasibility. By adjusting parameters such as search limit and pruning thresholds, the system can achieve near-optimal solutions with reduced computational burden, effectively tuning the balance between solution quality and computational resources required.