The demands for deterministic network services are getting significant. IEEE Time sensitive networking (TSN) task group has aimed for deterministic services through IEEE 802 networks. TSN's solutions rely on the time-synchronization and the slot scheduling coordinated throughout a network. The former requires costly hardware implementations for accuracy and scalability. The latter requires a schedulability analysis, which is proven to be NP-complete and not scalable. For modest-to-large scale dynamic networks, a more flexible solution is necessary. We propose a scalable framework that guarantees any desired jitter upper bound, including zero jitter. The framework is composed of; packets with relative time-stamps, a network that guarantees latency bounds, and buffers at the network egress. The latency bounds of a network can be of any value. The buffer holds packets for predetermined intervals to reproduce arbitrary inter-arrival intervals to inter-departure intervals, based on the network's latency bound. The idea behind the proposed framework is to minimize the latency first with work-conserving schedulers, and then adjust the jitter at the egress of a network. We argue that this is simpler than minimizing latency and jitter at the same time. A direct benefit of such simplicity is its scalability. We examine the network solutions that guarantee latency bound, such as the DiffServ and the Flow aggregate-interleave regulator (FAIR), as the candidates for the component in the framework. We show with a case study and simulations that the proposed framework can guarantee smaller latency bounds, with zero jitter, than those of the TSN's synchronous approach.

Photoplethysmography (PPG) is a simple and cost-efficient technique that effectively measures cardiovascular response by detecting blood volume changes in a noninvasive manner. A practical challenge in the use of PPGs in real-world applications is noise reduction. PPG signals are likely to be compromised by various types of noise, such as scattering or motion artifacts, and removing such compounding noises using a monotonous method is not easy. To this end, this paper proposes a neural PPG denoiser that can robustly remove multiple types of noise from a PPG signal. By casting the noise reduction problem into a signal restoration approach, we aim to achieve a solid performance in the reduction of different noise types using a single neural denoiser built upon transformer-based deep generative models. Using this proposed method, we conducted the experiments on the noise reduction of a PPG signal synthetically contaminated with five types of noise. Following this, we performed a comparative study using six different noise reduction algorithms, each of which is known to be the best model for each noise. Evaluation results of the peak signal-to-noise ratio (PSNR) show that the neural PPG denoiser is superior in three out of five noise types to the performance of conventional noise reduction algorithms. The salt-and-pepper noise type showed the best performance, with the PSNR of the neural PPG denoiser being 36.6080, and the PSNRs of the other methods were 19.8160 and 32.8234. The Poisson noise type performed the worst, showing a PSNR of 33.0090; the PSNRs of other methods were 35.1822 and 33.4795, respectively. Thereafter, an experiment to recover a signal synthesized with two or more of the five noise types was conducted. When the number of mixed noises was two, three, four, and five, the PSNRs were 29.2759, 27.8759, 26.5608, and 25.9402, respectively. Finally, an experiment to recover motion artifacts was also conducted. The synthesized motion artifact signal was created by synthesizing only a certain ratio of the total signal length. As a result of the motion artifact signal restoration, the PSNRs were 25.2872, 22.8240, 21.2901, and 19.9577 at 30%, 50%, 70%, and 90% motion artifact ratios, respectively. In the three experiments conducted, the neural PPG denoiser showed that various types of noise were effectively removed. This proposal contributes to the universal denoising of continuous PPG signals and can be further expanded to denoise continuous signals in the general domain.

In this study, we propose a scalable framework that guarantees both latency and jitter bounds in large networks, including the Internet. The framework is composed of two parts: a latency-guaranteeing network and a jitter-guaranteeing end system. For latency bounds, we suggest regulators per class per input–output port pair of the DiffServ-type relay nodes. For jitter bounds, based on the guaranteed latency bounds, we suggest time-stamping and buffers at the network egress edge. The framework does not require network-wide time synchronization, frequency synchronization, flow state maintenance, or flow-level queuing/scheduling. Therefore, the complexity does not increase as the number of flows or network size increases. Moreover, the framework is based on the DiffServ architecture; therefore, it requires minimal modification to the current Internet. We demonstrate that the proposed regulators can achieve latency bounds comparable to the IEEE asynchronous traffic shaping technique. We prove that the jitter is bounded even with realistic limitations such as buffers without cut-through capability. We also prove that in the presence of clock drift, the jitter can still be upper bounded with a suggested compensation algorithm. We demonstrate through experiments on simple programmable microcontrollers that the jitter upper bound can be within a few tens of microseconds, even in a realistic situation with store-and-forward buffers, clock drift, and random network delays.

Adopting reinforcement learning in the network scheduling area is getting more attention than ever because of its flexibility in adapting to the dynamic changes of network traffic and network status. In this study, a timeslot scheduling algorithm for traffic, with similar requirements but different priorities, is designed using a double deep q-network (DDQN), a reinforcement learning algorithm. To evaluate the behavior of the DDQN agent, a reward function is defined based on the difference between the estimated delay and the deadline of packets transmitted at the timeslot, and on the priority of packets. The simulation showed that the designed scheduling algorithm performs better than existing algorithms, such as the strict priority (SP) or weighted round robin (WRR) scheduler, in the sense that more packets arrived within the deadline. By using the proposed DDQN-based scheduler, it is expected that autonomous network scheduling can be realized in upcoming frameworks, such as time-sensitive or deterministic networking.

We consider guaranteeing end-to-end (E2E) latency bounds to flows in a network. It is desirable that flows are isolated from other flows. The bursts from other flows or the network utilization level should not affect a flow’s latency bound. The fair queuing technique, which includes Packetized Generalized Processor Sharing (PGPS) and Virtual Clock (VC), is based on the concept of ideal packet service completion time called the Finish Time (FT). The fair queuing is known to provide the best flow isolation but suffers from the fact that it has to maintain the flow states. Alternative schemes were suggested, in which the entrance node in a network generates FT for a packet and records it in the packet with other necessary information. Subsequent nodes, based on these records, decide the service eligibility and service order of the packets. A packet is served only when it is eligible, thus the system is non-work conserving. In this paper, a simpler framework for deriving such FTs in core nodes is presented, in which initial FT is updated by adding a delay factor that is a function of parameters of nodes and flow. The proposed scheduler is work conserving and has the property that, for a certain choice of the delay factor, the expression for E2E latency bound can be found. This E2E latency bound function is the same as that of a network with stateful fair queuing schedulers in all the nodes. Moreover, utilizing the fact that the service order of packets passing through the same path can be unaltered in the middle of the path, we also present a FIFO-based architecture whose performance is similar to that of a sorted queue-based architecture. The extensive simulations prove that the proposed framework shows an ideal flow isolation performance over a wide range of delay factor values, with superior scalability.

In services such as metaverse, which should provide a constant quality of service (QoS) regardless of the user’s physical location, the end-to-end (E2E) latency must be fairly distributed over any flow in the network. To this end, we propose a reinforcement learning (RL)-based scheduler for minimizing the maximum network E2E latency. The RL model used the double deep Q-network (DDQN) with the prioritized experience replay (PER). In order to see the performance change according to the type of RL agent, we implemented a single-agent environment where the controller is an agent and a multi-agent environment where each node is an agent. Since the agents were unable to identify E2E latencies in the multi-agent environment, the state and reward were formulated using the estimated E2E latencies. To precisely evaluate the RL-based scheduler, we set out benchmark algorithms to compare with which a network-arrival-time-based heuristic algorithm (NAT-HA) and a maximum-estimated-delay-based heuristic algorithm (MED-HA). The RL-based scheduler, first-in-first-out (FIFO), round-robin (RR), NAT-HA, and MED-HA were compared through large-scale simulations on four network topologies. The simulation results in fixed-packet generation scenarios showed that our proposal, the RL-based scheduler, achieved the minimization of maximum E2E latency in all the topologies. In other scenarios with random flow generation, the RL-based scheduler and MED-HA showed the lowest maximum E2E latency for all topologies. Depending on the topology, the maximum E2E latency of NAT-HA was equal to or larger than that of the RL-based scheduler. In terms of fairness, the RL-based scheduler showed a higher level of fairness than that of FIFO and RR. NAT-HA had similar or lower fairness than the RL-based scheduler depending on the topology, and MED-HA had the same level of fairness as the RL-based scheduler.