Impact of Delays on Constrained Online Convex Optimization

Cao, Xuanyu; Zhang, Junshan; Poor, H. Vincent

doi:10.1109/ieeeconf44664.2019.9048958

Cited by 6 publications

(21 citation statements)

References 7 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Delayadaptive online gradient descent algorithms were proposed in [14] and [15] to accommodate adversarial feedback delay. The impact of multi-slot feedback delay on OCO with longterm constraints was considered only in [16]. However, [16] only studied the static regret, which may not be an attractive performance metric for an inherently time-varying system.…”

Section: Related Workmentioning

confidence: 99%

“…The impact of multi-slot feedback delay on OCO with longterm constraints was considered only in [16]. However, [16] only studied the static regret, which may not be an attractive performance metric for an inherently time-varying system. Furthermore, the constraint violation bound provided in [16] was no less than O(T 3 4 ) even for time-invariant constraints, inherited from the saddle-point-typed algorithm [7].…”

Section: Related Workmentioning

confidence: 99%

“…However, [16] only studied the static regret, which may not be an attractive performance metric for an inherently time-varying system. Furthermore, the constraint violation bound provided in [16] was no less than O(T 3 4 ) even for time-invariant constraints, inherited from the saddle-point-typed algorithm [7]. Thus, the impact of multi-slot feedback delay on constrained OCO is still not well understood.…”

Section: Related Workmentioning

confidence: 99%

“…Therefore, [13] initiated a study on OCO with multi-slot feedback delay. Additional delay-adaptive OCO algorithms were proposed in [14] and [15], while a static regret bound was derived in recent work [16] for time-varying constrained OCO with multi-slot feedback delay.…”

Section: Introductionmentioning

confidence: 99%

“…• We apply DTC-OCO to a general network resource allocation problem, which arises in many systems such as data networks and cloud computing. Simulation based on an example cloud computing system demonstrates that DTC-OCO is tolerant to feedback delay and substantially outperforms the known best alternative from [16]. The rest of this paper is organized as follows.…”

Section: Introductionmentioning

confidence: 99%

See 4 more Smart Citations

Delay-Tolerant Constrained OCO with Application to Network Resource Allocation

Wang,

Liang,

Dong

et al. 2021

Preprint

View full text Add to dashboard Cite

We consider online convex optimization (OCO) with multi-slot feedback delay, where an agent makes a sequence of online decisions to minimize the accumulation of time-varying convex loss functions, subject to short-term and long-term constraints that are possibly time-varying. The current convex loss function and the long-term constraint function are revealed to the agent only after the decision is made, and they may be delayed for multiple time slots. Existing work on OCO under this general setting has focused on the static regret, which measures the gap of losses between the online decision sequence and an offline benchmark that is fixed over time. In this work, we consider both the static regret and the more practically meaningful dynamic regret, where the benchmark is a time-varying sequence of per-slot optimizers. We propose an efficient algorithm, termed Delay-Tolerant Constrained-OCO (DTC-OCO), which uses a novel constraint penalty with double regularization to tackle the asynchrony between information feedback and decision updates. We derive upper bounds on its dynamic regret, static regret, and constraint violation, proving them to be sublinear under mild conditions. We further apply DTC-OCO to a general network resource allocation problem, which arises in many systems such as data networks and cloud computing. Simulation results demonstrate substantial performance gain of DTC-OCO over the known best alternative.

show abstract

Section: Related Workmentioning

confidence: 99%