High speed and efficient accelerators such as neuromorphic photonics are becoming increasingly popular, as they can significantly improve computation speed and energy efficiency of matrix-based calculations, leading to femtojoule per MAC efficiency. However, deploying existing DL models on such platforms is not trivial, as a wide range of photonic neural network (PNN) architectures relies on incoherent setups and power-adding operational schemes that cannot natively represent negative quantities. This results in additional hardware complexity that increases cost and reduces energy efficiency. To overcome this, we can train non-negative neural networks and potentially exploit the full range of incoherent neuromorphic photonic capabilities. However, the existing non-negative training approaches cannot achieve the same level of accuracy as their regular counterparts. To this end, we introduce a methodology to obtain the non-negative isomorphic equivalents of regular artificial neural networks (ANNs) that meet the requirements of neuromorphic hardware as it allows one to perform inference with non-negative only parameters. Furthermore, we also introduce a sign-preserving optimization approach that enables the training of such isomorphic networks in a non-negative manner. The source code implementation of the proposed numerical framework is publicly available at: https://github.com/eakirtas/nn_isomorphics_24 .

Abstract Photonic neural networks have emerged as a promising solution to overcome limitations of traditional hardware for neuromorphic computations, offering advantages in bandwidth, latency, and power efficiency. However, their performance is constrained by the limited precision of analog photonic computing, which is affected by inherent noise sources such as thermal and shot noise, and distortions. These effects degrade the photonic neural network performance, reducing the bit resolution achievable in photonic hardware typically to 2–4 bits. Traditional quantization strategies fail to account for these noise contributions, resulting in a substantial accuracy loss during inference. This paper introduces an adaptive quantization method called Adaptive-Quantization Photonic-Aware Neural Network (AQ-PANN) to address the challenges posed by different noise sources in analog photonic hardware. The proposed method uses a learnable step size quantization scheme to achieve high accuracy and stability under varying noise levels, introducing a scheme that unifies quantization step adaptation with noise injection exactly where photonic distortions arise. This design incurs only a minor training-time overhead, as it involves learning a small number of per-layer quantization step sizes and does not affect inference. Experimental evaluations on three commonly used test datasets (MNIST, SVHN, and CIFAR-10) with different bit resolutions show the robustness of AQ-PANN. On MNIST, an accuracy drop of only 2% was observed from low to high noise levels in a 4-bit configuration, while traditional DoReFa quantization suffered a 29% drop. For the SVHN dataset, AQ-PANN obtained a mean accuracy of 92% under high noise with 4-bit quantization, outperforming DoReFa by over 45%. On CIFAR-10, AQ-PANN maintained close to 60% accuracy under high noise in the 4-bit configuration, whereas DoReFa and PACT both collapsed below 40%. These results highlight the effectiveness of AQ-PANN in sustaining model performance across different noise intensities, enabling practical photonic neural network deployment.

The ePhos project addresses the need to inspire the next generation in photonics, a transformative field shaping various aspects of our lives. Comprising a community of practice and a mobile augmented reality (AR) application named ePhosAR. These tools provide free and accessible photonics educational material for diverse audiences. The community of practice fosters collaborative learning through educational material, while the ePhosAR app, leveraging AR technology, transforms theoretical concepts into engaging experiences. The app is also accompanied by a board game for interactive learning. The dissemination efforts conducted at schools and universities received quite positive feedback, affirming the project's potential. The objectives of ePhos tools are to make photonics education easy, enjoyable, and accessible. Future work involves enriching and researching the tools' effectiveness in photonics education.

In this paper, we jointly study the real-time user association, traffic routing and x-Network Function (xNF) placement problem in an integrated Terrestrial and Satellite 6G Network (TN-SN), with the aim of maximizing the network energy efficiency and user acceptance ratio. We formulate the aforementioned problem as a Mixed Integer Linear Program (MILP), considering various capacity, power and flow conservation constraints for the integrated network, while also meeting the specific service requirements of each user. To tackle the increased complexity of the optimal solution, we also develop an efficient heuristic (named as TERA). Through extensive simulations, TERA demonstrates a notable superiority in energy efficiency compared to the current State-of-the-Art (SoA), achieving up to 85 % of the optimal with up to 87 % lower execution time even under challenging traffic load conditions.

We present an optimized Medium-Transparent MAC (oMT-MAC) protocol for Analog-RoF Fiber-Wireless 5G and beyond X-haul networks, capable of generating an optimum transmission schedule, which boosts the protocol's efficiency by utilizing the full delay budget of higher priority flows. Results show that oMT-MAC's optimization model can reduce delays by up to 20% for low-priority flows while maintaining the 5G Fronthaul and URLLC KPIs.

We propose an enhanced Medium-Transparent MAC (eMT-MAC) protocol for Analog-RoF Fiber-Wireless 5G and beyond X-haul networks, capable of handling variable packet sizes and wavelength sharing among multiple Remote Antenna Units (RAUs), thus increasing the protocol's efficiency for low-payload-length services such as URLLC/mMTC, while adding support for Point-to-Point links. Results show that eMT-MAC achieves high protocol efficiency, while also supporting wavelength sharing, and meets Low-layer split fronthaul specifications even for wavelength switching delays up to$12\mu\mathrm{s}$.

Abstract The energy demanding nature of deep learning (DL) has fueled the immense attention for neuromorphic architectures due to their ability to operate in a very high frequencies in a very low energy consumption. To this end, neuromorphic photonics are among the most promising research directions, since they are able to achieve femtojoule per MAC efficiency. Although electrooptical substances provide a fast and efficient platform for DL, they also introduce various noise sources that impact the effective bit resolution, introducing new challenges to DL quantization. In this work, we propose a quantization-aware training method that gradually performs bit reduction to layers in a mixed-precision manner, enabling us to operate lower-precision networks during deployment and further increase the computational rate of the developed accelerators while keeping the energy consumption low. Exploiting the observation that intermediate layers have lower-precision requirements, we propose to gradually reduce layers’ bit resolutions, by normally distributing the reduction probability of each layer. We experimentally demonstrate the advantages of mixed-precision quantization in both performance and inference time. Furthermore, we experimentally evaluate the proposed method in different tasks, architectures, and photonic configurations, highlighting its immense capabilities to reduce the average bit resolution of DL models while significantly outperforming the evaluated baselines.

Conventional electronic memory hierarchies are intrinsically limited in their ability to overcome the memory wall due to scaling constraints. Optical caches and interconnects can mitigate these constraints, and enable processors to reach performance and energy efficiency unattainable by purely electronic means. However, the promised benefits cannot be realized through a simple replacement process; to reach its full potential, the architecture needs to be holistically redesigned. This article proposes Pho$, an opto-electronic memory hierarchy architecture for multicores. Pho$ replaces conventional core-private electronic caches with a large shared optical L1 built with optical SRAMs. The shared optical cache is supported by Pho$Net, a novel hybrid MWSR/R-SWMR optical NoC that provides low-latency and high-bandwidth communication between the electronic cores and the shared optical L1 at low optical loss. Pho$Net’s unique network arbitration protocol seamlessly co-arbitrates the request and reply sub-networks and facilitates cache requests and replies that optimize for the common case of cache hits. Through Pho$ we solve the problems that render previous designs impractical. Our results show that Pho$ achieves on average 1.41× performance speedup (3.89× max) and 31% lower energy-delay product (90% max) against conventional designs. Moreover, the Pho$Net optical NoC for core-cache communication consumes 70% less power compared to directly applying previously proposed optical NoC architectures.

We study the problem of downlink Coordinated Beamforming for dense fixed-wireless millimeter wave (mmWave) networks under a Centralized/Cloud Radio Access Network (C-RAN) setting. To compensate for the increased mmWave path loss, we consider directional transmissions via antenna array analog beamforming, which leads to adiscreteset of available beams. We apply the Lyapunov optimization framework to propose a throughput-optimal policy which explicitly accounts for stochastic traffic and channel fluctuations and dynamically performs joint Base Station-to-user association and analog beam selection. Our model makes minimal assumptions and considers realistic mmWave antenna radiation patterns, while it can be easily extended to include additional MAC and PHY layer controls. Since this flexibility comes at the cost of high computational complexity, we also propose two heuristic policies offering reduced complexity. Their performance is compared against a baseline Round-Robin (RR)-based policy, while a theoreticalworst-caseperformance bound is derived. Extensive simulations show the optimal Lyapunov policy achieving a stable throughput increase of more than 2Xper usercompared to the RR policy, while the proposed heuristics incur only a 20-30% performance penalty with respect to the optimal Lyapunov policy with 60X-900X computational time savings.

Network Function Virtualization (NFV) is expected to be a crucial enabler of Beyond 5G (B5G) networks, supporting new services with stringent requirements, while offering high flexibility and configurability. In this paper, we consider the joint problem of Virtual Network Function (VNF) placement alongside computational and communication resource allocation in a mobile network to minimize total power expenditure while satisfying the Service Function Chain (SFC), throughput and delay requirements. We explicitly account for the Access Network (AN) and formulate a general Mixed Integer Linear Program (MILP). Due to the high complexity of the latter, we propose an efficient heuristic, which is shown to significantly outperform the State-of-Art (SoA), while achieving up to 78% of the optimal energy efficiency with up to 742 times lower complexity.

Conventional electronic memory hierarchies are intrinsically limited in their ability to overcome the memory wall due to scaling constraints. Optical caches and interconnects can mitigate these constraints, and enable processors to reach performance and energy efficiency unattainable by purely electronic means. However, the promised benefits cannot be realized through a simple replacement process; to reach its full potential, the architecture needs to be holistically redesigned. This paper proposes $\text { Pho\$ }$, an opto-electronic memory hierarchy architecture for multicores. $\text { Pho\$ }$ replaces conventional coreprivate electronic caches with a large shared optical L1 built with optical SRAMs. A novel optical NoC provides low-latency and high-bandwidth communication between the electronic cores and the shared optical L1 at low optical loss. Our results show that $\text { Pho\$ }$ achieves on average $1.41\times$ performance speedup $(3.89 \times \max)$ and 31% lower energy-delay product (90% max) against conventional designs. Moreover, the optical NoC for core-cache communication consumes 70% less power compared to directly applying previously-proposed optical NoC architectures.

We demonstrate experimentally a bandwidth-reconfigurable mmWave Fiber Wireless (FiWi) fronthaul bus topology for spectrally efficient and flexibly reconfigurable 5G Centralized-Radio Access Networks (C-RAN). The proposed fronthaul architecture includes four 1 Gb/s Intermediate Frequency over Fiber (IFoF) channels that can be flexibly allocated among two in-series Reconfigurable Optical Add/Drop Multiplexer (ROADM) integrated nodes, supporting in total 8 V-band 32-element Phased Array Antenna (PAA) terminals. The ROADM was fabricated as an integrated photonic device exploiting the ultra-low loss Si3N4TriPleX waveguide integration platform and an architectural layout based on cascaded MZI interleavers. The device has flat top response of 32.5 GHz with a Free Spectral Range (FSR) of 100 GHz and fiber-to-fiber losses of 5 dB, while the V-band PAA supports analog RF beamsteering capabilities within a 90°-sector and 1m wireless distance. Each of the FiWi links carries a 250 MBd QAM16 waveform enabling a total of 1 Gb/s rate per end user beam, complying with the 5G Key Performance Indicator (KPI) user-rate requirement. Bandwidth-reconfigurability is experimentally demonstrated by selectively dropping channels either at the first or at the second ROADM node, allowing in this way the bandwidth allocation to be flexibly defined between two different network segments. Both uplink and downlink performance are experimentally validated for different ratios of bandwidth allocation among the two nodes, revealing Error Vector Magnitude (EVM) values that meet the respective 3GPP signal quality specifications. The two-stage FiWi IFoF/mmWave fronthaul bus topology, based on a miniaturized, integrated, low loss Si3N4ROADM and supporting high-capacity wireless beamsteering capability can form a promising roadmap towards flexible and reconfigurable 5G C-RAN architectures.

In order to meet the ever-increasing traffic demands, the combination of fiber and Millimeter Wave (mmWave) is expected to play a key role for 5G Centralized-Radio Access Networks (C-RANs). Due to the inefficiency of the Common Public Radio Interface for the Baseband Unit (BBU)-Remote Radio Head (RRH) communication, analog-Radio-over-Fiber (a-RoF) technology is considered a promising solution, mainly due to the RRH simplification and lower fronthaul requirements it imposes. In such mmWave a-RoF C-RANs, efficient Medium Transparent-Medium Access Control (MT-MAC) protocols are needed able to meet the challenging 5G requirements. To this end, in this paper, we propose a gated service MT-MAC protocol which authorizes each user to transmit the amount of data it requested. A detailed delay model is proposed, which is validated through simulations for different fiber lengths, network load conditions and number of available optical wavelengths. Moreover, the proposed protocol is compared with the state-of-the-art (SoA) and is shown to achieve up to 20 times higher throughput, 2 times lower delay with 100% lower blocking probability and 5 times higher data wavelength utilization, while being able to adapt to varying network traffic conditions. Our proposal also attains sub-ms latency in most cases, constituting it a promising candidate for 5G mmWave a-RoF C-RANs.

Photonic Deep Learning (DL) accelerators are among the most promising approaches for providing fast and energy efficient neural network implementations for several applications. However, photonic accelerators require using different activation functions compared to those typically used in DL. This renders the training process especially difficult to tune, often requiring several trials just for selecting the appropriate initialization hyper-parameters for the network. This process becomes even more difficult for recurrent networks, where exploding gradient phenomena can further destabilize the training process. In this paper, we propose an adaptive data-driven initialization approach for recurrent photonic neural networks. The proposed method is activation-agnostic, while it takes into account the actual distribution of the data used to train the network, overcoming a number of significant limitations of existing approaches. The proposed method is simple and easy to implement, yet it leads to significant improvements in the performance of DL models, as it was experimentally demonstrated using two large-scale challenging time-series datasets.

5G New Radio's (NR) spectrum expansion towards higher bands, although critical towards achieving the envisioned 5G capacity requirements, creates the need for installing a very large number of Access Points (APs), which asserts tremendous capital burden on the Mobile Network Operators. Current centralization solutions such as the Cloud Radio Access Network (C-RAN) alleviate partially the costs of densification by moving the majority of radio processing functionalities from the Remote Radio Heads (RRHs) to the central Base Band Unit (BBU), but still require very high-speed Point-to-Point links between the BBU and each RRH mainly due to the digitized Common Public Radio Interface (CPRI) that is excessively inefficient for hauling broadband signals. In this article, we present a novel architecture that employs an analog converged Fiber-Wireless scheme in order to create a very spectrally efficient Point-to-Multipoint network capable of interconnecting a large number of APs, while allowing compatibility with mature Ethernet-based low-cost equipment. Preliminary simulation results show very low end-to-end Ethernet packet delay, well below eCPRI's 100 μs mark, even for fiber lengths up to 10 km, indicating the suitability of our solution for employment in 5G NR large-scale fronthaul networks.

Photonic neuromorphic hardware can provide significant performance benefits for Deep Learning (DL) applications by accelerating and reducing the energy requirements of DL models. However, photonic neuromorphic architectures employ different activation elements than those traditionally used in DL, slowing down the convergence of the training process for such architectures. An initialization scheme that can be used to efficiently train deep photonic networks that employ quadratic sinusoidal activation functions is proposed in this paper. The proposed initialization scheme can overcome these limitations, leading to faster and more stable training of deep photonic neural networks. The ability of the proposed method to improve the convergence of the training process is experimentally demonstrated using two different DL architectures and two datasets.

Fifth Generation (5G) Cloud-Radio Access Networks (C-RANs) are about to exploit both optical and Millimeter Wave (mmWave) technology to meet the ever-increasing traffic demands. In this new type of converged Fiber-Wireless (FiWi) systems efficient Medium Transparent-Medium Access Control (MT-MAC) protocols should be designed, able to satisfy the very strict 5G service requirements. To this end, in this paper, we propose an MT-MAC protocol for mmWave Analog Radio-over-Fiber (A-RoF) C-RANs, which employs gated service, according to which users are granted transmission windows equal to the number of bytes contained in their buffer. An analytical model is also proposed for the mean packet delay, which is verified by means of simulation for different fiber length values, network load conditions and optical capacity values. Our results not only prove the accuracy of the proposed model but also the suitability of the proposed MT-MAC protocol to meet the sub-ms delay challenge of latency-critical 5G network requirements.

The upcoming generation of mobile networks is expected to serve numerous mobile users with high quality-of-service (QoS) demands, requiring high-capacity fronthaul. As the provision of fiber connections directly to the end users is not cost-efficient, the integrated fiber wireless (FiWi) fronthaul design based on wireless networking and passive optical networks (PONs) has been proposed. The FiWi design involves modern networking technologies that can accommodate the need for data rates in the Gb/s scale and low delay, such as the wavelength division multiplexing (WDM) in the optical domain and the multiple input multiple output (MIMO) communication over millimeter wave (mmWave) spectrum in the wireless domain. The co-existence of two network types requires resource management in a medium transparent manner, i.e., the sharing of the bandwidth in the wireless domain should allow the organization of the data packets in optical frames. As the traffic circulating in the FiWi fronthaul involves packets of different priorities, i.e., different QoS classes, the resource management scheme should support QoS differentiation. To this end, we propose a resource management scheme for FiWi fronthaul and we extensively study its performance in terms of experienced delay and throughput. Our simulation results demonstrate that the proposed scheme significantly reduces the delay of the high priority class.

We demonstrate for the first time an analytical model for computing the end to end packet delay of an Optical/Wireless 60GHz Radio-over-Fiber (RoF) network operating under the Medium-Transparent MAC (MT-MAC) protocol. The model takes into account contention both at the optical and the wireless layer, effectively incorporating the MT-MAC mechanism for seamless and dynamic capacity allocation over both optical and wireless transmission media. Based on this model, we provide an extensive delay performance analysis of the Medium Transparent MAC protocol for various optical capacity availability scenarios, varying load conditions and optical network fiber lengths. The theoretical results are found to be in good agreement with respective simulation-based findings, confirming that the employment of Medium Transparent MAC protocols can allow for efficient incorporation of the MT-MAC scheme into the upcoming era of 5G mm-wave small-cell networks.

We demonstrate for the first time an analytical model for computing the saturation throughput of a Medium-Transparent MAC protocol in 60 GHz Radio-over-Fiber networks, assuming a finite number of terminals and ideal channel conditions. The model takes into account contention both at the optical and the wireless layer, effectively incorporating the Medium-Transparent MAC mechanism for seamless and dynamic capacity allocation over both optical and wireless transmission media. Based on this model, we provide an extensive saturation throughput performance analysis of the Medium Transparent MAC protocol for various optical capacity availability scenarios and different numbers of wireless users. The theoretical results are found to be in good agreement with respective simulation-based findings, confirming that the employment of Medium Transparent MAC protocols can allow for efficient extended LAN operation of 60 Radio-over-Fiber networks.

We demonstrate a handover scheme for supporting high end-user mobility in a 60 GHz broadband picocellular radio-over-fiber network. The proposed scheme relies on a novel moving extended cell (MEC) concept which is based on user-centric virtual groups of adjacent cells that transmit the same data content to the user. The MEC concept utilizes a mechanism for restructuring the virtual multi-cell area according to the user's mobility pattern, so that a virtual antenna group moves together with the mobile user. A mathematical model is developed and performance evaluation is also conducted and presented, demonstrating zero packet loss and call dropping probability values in high-rate wireless services for a broad range of mobile speeds up to 40 m/sec and independently of the fiber link distances.

All-optical label swapping (AOLS) has been introduced as a new routing paradigm based on Multi-protocol Label Swapping (MPLS), aiming at bridging the gap between fiber and router packet forwarding capacity. To that end, AOLS routers should handle channel capacities of 40 Gb/s or higher, operating with packets of arbitrary lengths that are potentially generated by packet aggregation at the ingress node. In this communication we report on the implementation of all-optical 40 Gb/s 3R regeneration and label/payload separation sub-systems, operating with different-length packet-mode traffic and how these sub-systems find application in future AOLS networks. Both circuits were implemented by using optically-controlled integrated Mach-Zehnder Interferometric switches as the basic switching elements. The 40 Gb/s 3R regenerator showed high jitter tolerance and error-free operation exhibiting negative power penalty of 2.5 dB. Successful separation of the label from the payload was achieved with 40 Gb/s short packets of different payload length, while error-free operation with 3 dB penalty is demonstrated.