Skip to content

2 papers indexed here

We haven’t gathered this author’s papers yet. Follow them and we’ll fetch their work.

Not the right person? Other researchers publish under this name.

Conference Aug 2026

Elastic TDMA Overlay for Dense Wi-Fi Networks

Dense Wi-Fi networks can lose efficiency when many uplink stations contend at the same time. This paper presents an access-policy study of an AP-managed elastic TDMA-style scheduled-access overlay for dense WLANs. The intended policy reserves protected airtime for active hybrid-capable stations while leaving fallback CSMA/CA available for regular stations and new reservation requests. We evaluate the policy with a policy-level ns-3 model layered over stock Wi-Fi behavior. The model uses scheduled queue release and NAV-like regular-station deferral to isolate access-policy behavior before full MAC integration. In a matched 16-station high-load campaign, a mixed 8 hybrid / 8 regular configuration with NAV-like deferral improves throughput by 6.51%, delivery by 5.74 percentage points, Wi-Fi TX failures by $\mathbf{6 0 . 3 4 \%}$, and retransmissions by $\mathbf{2 8 . 1 8 \%}$ relative to all-regular standard Wi-Fi. Paired five-run differences give preliminary statistical support for the protected mixed case, including throughput gains of $1.175 \pm 0.372$ Mbps and delivery gains of $5.743 \pm 1.820$ percentage points. A NAV-like-off stress case degrades, representing missed or unenforced protection. A fixed 8H/8R timing sweep shows that protected airtime must be sized carefully: 5 ms is best among the tested block durations, while oversized protected blocks starve fallback contention. These results support AP-managed elastic TDMA-style access policy and motivate future MAC-integrated validation.

Reuven Mueller, Ying Xie · 0 citations
Conference Open access 2026

AgentSearch: Learning Efficient Agentic Workflows via Deliver Tree Search

This work introduces AgentSearch, a cost-aware Monte Carlo Tree Search (MCTS) framework that constructs agentic workflows through deliberative lookahead search and attains single-episode success while reducing computational costs by up to 47%, thereby eliminating the trial-and-error exploration required by previous adaptive methods.

D. Attota, Ying Xie · 0 citations