Target Wake Time Scheduling for Time-sensitive and Energy-efficient Wi-Fi Networks

TL;DR

The paper addresses the challenge of achieving time-sensitive, energy-efficient communication in Wi-Fi by leveraging Target Wake Time (TWT) to schedule non-overlapping transmission windows. It formulates the TWT Acceptance and Scheduling Problem (TASP) as a mixed-integer quadratic constrained program that jointly considers throughput, energy, and Age of Information (AoI), proving NP-hardness, and introduces TASPER, a graph-based heuristic that searches high-value scheduling paths within a controllable neighborhood. The authors extend an ns-3-based TWT simulator (ns-3-twt) with AoI deadlines and energy models, and validate TASPER through extensive simulations and a real 10-node IIoT testbed, showing up to ~25% lower rejection cost and substantial energy savings compared with baselines such as ShortestFirst and HSA. The work demonstrates that integrated TWT scheduling can deliver deterministic, energy-efficient, time-sensitive Wi-Fi performance suitable for Industrial IoT, with practical validation on commodity hardware and clear directions for future OFDMA-enabled extensions.

Abstract

Time Sensitive Networking (TSN) is fundamental for the reliable, low-latency networks that will enable the Industrial Internet of Things (IIoT). Wi-Fi has historically been considered unfit for TSN, as channel contention and collisions prevent deterministic transmission delays. However, this issue can be overcome by using Target Wake Time (TWT), which enables the access point to instruct Wi-Fi stations to wake up and transmit in non-overlapping TWT Service Periods (SPs), and sleep in the remaining time. In this paper, we first formulate the TWT Acceptance and Scheduling Problem (TASP), with the objective to schedule TWT SPs that maximize traffic throughput and energy efficiency while respecting Age of Information (AoI) constraints. Then, due to TASP being NP-hard, we propose the TASP Efficient Resolver (TASPER), a heuristic strategy to find near-optimal solutions efficiently. Using a TWT simulator based on ns-3, we compare TASPER to several baselines, including HSA, a state-of-the-art solution originally designed for WirelessHART networks. We demonstrate that TASPER obtains up to 24.97% lower mean transmission rejection cost and saves up to 14.86% more energy compared to the leading baseline, ShortestFirst, in a challenging, large-scale scenario. Additionally, when compared to HSA, TASPER also reduces the energy consumption by 34% and reduces the mean rejection cost by 26%. Furthermore, we validate TASPER on our IIoT testbed, which comprises 10 commercial TWT-compatible stations, observing that our solution admits more transmissions than the best baseline strategy, without violating any AoI deadline.

Figures (12)

• Figure 1: Example of the temporal evolution of STA's power states with (a) and without (b) TWT, over a period of 32 ms. The total energy consumed during the time period is reported at the bottom right. The example includes 8 STAs, each with a 1,500-byte UDP packet to transmit; the considered STA starts transmitting at time 10 ms.
• Figure 2: Comparison of the average STA transmission delay measured with and without TWT, when used to avoid channel contention. Error bars show the 95% confidence intervals.
• Figure 3: TASP example (a) and its solution using a FIFO strategy (b) and a strategy maximizing the number of transmissions (c). The blue bars indicate the transmission times, which have to start and end inside the white boxes to meet the traffic deadlines.
• Figure 4: System model overview. At each beacon interval, STAs first request traffic scheduling from TASPER, and then follow its scheduling decisions.
• Figure 5: Example of the TASPER graph, referring to the scheduling problem in Fig. \ref{['fig:naivevsoptimum']}. For clarity, the outbound edges have the same color as the vertex they exit while their arrow has the same color as the inbound vertex. Notice how the graph is not fully connected: e.g., there is no edge from TX 3 to TX 4, since, if TX 3 is scheduled, its end time would exceed the latest possible start time of TX 4. The solid thick lines, traversing vertices $\alpha$, $2$, $7$, $5$, $8$, $1$, $6$, and $\omega$, mark the best path, i.e., the one identified by the Optimum solution in Fig. \ref{['fig:naivevsoptimum']}. Finally, the dotted thick lines identify the neighborhood of Tx $4$ with a neighborhood size $\eta{=}1$, including TXs 5 and 7.

Theorems & Definitions (2)

• Theorem 1
• proof