Wireless networks are being applied in various cyber-physical systems and posed to support

mission-critical cyber-physical systems applications. When those applications require reliable and

low-latency wireless communication, ensuring predictable per-packet communication reliability is a

basis. Due to co-channel interference and wireless channel dynamics (e.g. multi-path fading), however,

wireless communication is inherently dynamic and subject to complex uncertainties. Power

control and MAC-layer scheduling are two enablers. In this dissertation, cross-layer optimization

of joint power control and scheduling for ensuring predictable reliability has been studied. With an

emphasis on distributed approaches, we propose a general framework and additionally a distributed

algorithm in static networks to address small channel variations and satisfy the requirements on

receiver-side signal-to-interference-plus-noise-ratio (SINR). Moreover, toward addressing reliability

in the settings of large-scale channel dynamics, we conduct an analysis of the strategy of joint

scheduling and power control and demonstrate the challenges.

First, a general framework for distributed power control is considered. Given a set of links

subject to co-channel interference and channel dynamics, the goal is to adjust each link's transmission

power on-the-fly so that all the links' instantaneous packet delivery ratio requirements

can be satised. By adopting the SINR high-delity model, this problem can be formulated as

a Linear Programming problem. Furthermore, Perron-Frobenius theory indicates the characteristic

of infeasibility, which means that not all links can nd a transmission power to meet all the

SINR requirements. This nding provides a theoretical foundation for the Physical-Ratio-K (PRK)

model. We build our framework based on the PRK model and NAMA scheduling. In the proposed

framework, we dene the optimal K as a measurement for feasibility. Transmission power and

scheduling will be adjusted by K and achieve near-optimal performance in terms of reliability and

concurrency.

Second, we propose a distributed power control and scheduling algorithm for mission-critical

Internet-of-Things (IoT) communications. Existing solutions are mostly based on heuristic algorithms

or asymptotic analysis of network performance, and there lack eld-deployable algorithms

for ensuring predictable communication reliability. When IoT systems are mostly static or low mobility,

we model the wireless channel with small channel variations. For this setting, our approach

adopts the framework mentioned above and employs feedback control for online K adaptation and

transmission power update. At each time instant, each sender will run NAMA scheduling to determine

if it can obtain channel access or not. When each sender gets the channel access and sends a

packet, its receiver will measure the current SINR and calculate the scheduling K and transmission

power for the next time slot according to current K, transmission power and SINR. This adaptive

distributed approach has demonstrated a signicant improvement compared to state-of-the-art

technique. The proposed algorithm is expected to serve as a foundation for distributed scheduling

and power control as the penetration of IoT applications expands to levels at which both the

network capacity and communication reliability become critical.

Finally, we address the challenges of power control and scheduling in the presence of large-scale

channel dynamics. Distributed approaches generally require time to converge, and this becomes a

major issue in large-scale dynamics where channel may change faster than the convergence time

of algorithms. We dene the cumulative interference factor as a measurement of impact of a single

link's interference. We examine the characteristic of the interference matrix and propose that

scheduling with close-by links silent will be still an ecient way of constructing a set of links

whose required reliability is feasible with proper transmission power control even in the situation of

large-scale channel dynamics. Given that scheduling alone is unable to ensure predictable communication

reliability while ensuring high throughput and addressing fast-varying channel dynamics,

we demonstrate how power control can help improve both reliability at each time instant and

throughput in the long-term. Collectively, these ndings provide insight into the cross-layer design

of joint scheduling and power control for ensuring predictable per-packet reliability in the presence

of wireless network dynamics and uncertainties.