QDAP: A Queue-Driven Adaptive Protocol

QDAP (Queue-Driven Adaptive Protocol) is an open-source adaptive transport protocol engineered for high-reliability, low-latency communication in dynamic and resource-constrained environments. While traditional protocols such as HTTP, gRPC, UDP, and MQTT are optimised for stable infrastructure, they exhibit critical failures under unreliable channels, high mobility, and strict timing requirements — conditions common in vehicular networks, UAV communications, industrial IoT, and real-time distributed systems. QDAP addresses these limitations through four tightly integrated mechanisms: (1) a QFT-inspired log-linear softmax scheduler processing 2.6 million priority decisions per second with deterministic emergency preemption; (2) an exact-binomial Forward Error Correction (FEC) model achieving sub-0.1% residual loss at 30% channel packet-error rate; (3) a delta compression layer reducing telemetry payload by up to 74.4%; and (4) a GhostSession 0-RTT reconnection mechanism eliminating handshake latency on link recovery. We evaluate QDAP against five baselines (DSRC/802.11p, IEEE 802.11bd, C-V2X Mode 4, UDP, MQTT) across 450 Monte Carlo scenarios. In WAN experiments, QDAP achieves 96x higher throughput and 65x lower p99 latency under crisis conditions. In V2X simulation, QDAP maintains message delivery above 99% across all channel conditions while keeping end-to-end latency below 10 ms at densities up to 100 nodes. The open-source implementation additionally includes an Ephemeral Message Layer (EML) for protocol-level TTL-aware framing and an Event-Driven Selective Persistence (ESP) layer for blockchain-integrated anomaly logging — both actively developed and available in the repository.