Neural Code Translation With LIF Neuron Microcircuits.

Mobile Adhoc Networks (MANETs) typically employ with the aid of new technology to increase Quality-of-Service (QoS) when forwarding multiple data rates. This kind of network causes high forwarding delays and improper data transfer rates because of the changes in the node’s vicinity. Although an optimized routing technique to transfer energy has been used to lessen the delay and improve the throughput by assigning a proper data rate, it does not consider the objective of minimizing the energy use, which results in less network lifetime. The goal of the proposed work is to minimize the energy depletion in a MANET, which results in an extended Lifespan of the network. In this research paper, an Extended Life span and QSSM-ML routing algorithm is proposed, which minimizes energy use and enhances the network lifetime. First, an optimization problem is formulated with the purpose of increasing the network’s lifetime while limiting the energy utilization and stability of the path along with residual. Second, an adaptive policy is applied for the asymmetric distribution of energy at both origin and intermediate nodes. In order to achieve maximum network lifespan and minimal energy depletion, the optimization problem was framed when power usage is a constraint by allowing the network to make use of the leftover power. An asymmetric energy transmission strategy was also designed for the adaptive allocation of maximum transmission energy in the origin. This made the network lifespan extended with the help of reducing the node’s energy use for broadcasting the data from the origin to the target. Moreover, the node’s energy use during packet forwarding is reduced to recover the network lifetime. The overall benefit of the proposed work is that it can achieve both minimal energy depletion and maximizes the lifetime of the network. Finally, the simulation findings reveal that the ELQSSM-ML algorithm accomplishes a better network performance than the classical algorithms.

Spiking neural networks (SNNs) underlie low-power, fault-tolerant information processing in the brain and could constitute a power-efficient alternative to conventional deep neural networks when implemented on suitable neuromorphic hardware accelerators. However, instantiating SNNs that solve complex computational tasks in-silico remains a significant challenge. Surrogate gradient (SG) techniques have emerged as a standard solution for training SNNs end-to-end. Still, their success depends on synaptic weight initialization, similar to conventional artificial neural networks (ANNs). Yet, unlike in the case of ANNs, it remains elusive what constitutes a good initial state for an SNN. Here, we develop a general initialization strategy for SNNs inspired by the fluctuation-driven regime commonly observed in the brain. Specifically, we derive practical solutions for data-dependent weight initialization that ensure fluctuation-driven firing in the widely used leaky integrate-and-fire neurons. We empirically show that SNNs initialized following our strategy exhibit superior learning performance when trained with SGs. These findings generalize across several datasets and SNN architectures, including fully connected, deep convolutional, recurrent, and more biologically plausible SNNs obeying Dale’s law. Thus fluctuation-driven initialization provides a practical, versatile, and easy-to-implement strategy for improving SNN training performance on diverse tasks in neuromorphic engineering and computational neuroscience.

Large-scale spiking neural networks (SNNs) have been used to successfully model complex neural circuits that explore various neural phenomena such as learning and memory, vision systems, auditory systems, neural oscillations, and many other important topics of neural function. Additionally, SNNs are particularly well-adapted to run on neuromorphic hardware as spiking events are often sparse, leading to a potentially large reduction in both bandwidth requirements and power usage. The inclusion of realistic plasticity equations, neural dynamics, and recurrent topologies has increased the descriptive power of SNNs but has also made the task of tuning these biologically realistic SNNs difficult. We present an automated parameter-tuning framework capable of tuning large-scale SNNs quickly and efficiently using evolutionary algorithms (EA) and off-the-shelf graphics processing units (GPUs).To test the feasibility of an automated parameter-tuning framework, our group used EAs to tune open parameters in SNNs running concurrently on a GPU. The SNNs were evolved to produce orientation-dependent stimulus responses similar to those found in simple cells of the primary visual cortex (V1) through the formation of self-organizing receptive fields (SORFs). The general evolutionary approach was as follows: A population of neural networks was created, each with a unique set of neural parameter values that defined overall behavior. Each SNN was then ranked based on a fitness value assigned by an objective function in which higher fitness values were given to SNNs that (a) reproduced responses observed in primate visual cortex, and (b) spanned the stimulus space, and (c) had sparse firing rates. The highest ranked individuals were selected, recombined, and mutated to form the offspring for the next generation. This process continued until a desired fitness was reached or until other termination conditions were met (Figure 1a).The automated parameter-tuning framework consisted of three software packages. The framework included: (a) the CARLsim SNN simulator [1], (2) the Evolving Objects (EO) computational framework [2], and (3) a parameter-tuning interface (PTI), developed by our group, to provide an interface between CARLsim and EO (See Figure 1b). The EO computational framework ran the evolutionary algorithm on the user-designated parameters of SNNs in CARLsim. The PTI allowed the objective function to be calculated independent of the EO computation framework. Parameter values were passed from the EO computation framework through the PTI to the SNN in CARLsim where the objective function is calculated. After the objective function was executed, the results were passed from the SNN in CARLsim through the PTI back to the EO computation framework for processing by the EA. With this approach, the fitness function calculation, which involved running each SNN in the population, could be run in parallel on the GPU while the remainder of EA calculations can be performed using the CPU (Figure 1b).A sample SNN with 4,104 neurons was tuned to respond with V1 simple cell-like tuning curves and produce SORFs. A performance analysis comparing the GPU-accelerated implementation to a single-threaded CPU implementation was carried out and showed that the GPU implementation could achieve a 65 times speedup over the CPU implementation. Additionally, the parameter value solutions found in the tuned SNN were stable and robust.The automated parameter-tuning framework presented here will be of use to both the computational neuroscience and neuromorphic engineering communities, making the process of constructing and tuning large-scale SNNs much quicker and easier.

Spiking neural networks (SNNs) have garnered interest due to their energy efficiency and superior effectiveness on neuromorphic chips compared with traditional artificial neural networks (ANNs). One of the mainstream approaches to implementing deep SNNs is the ANN-SNN conversion, which integrates the efficient training strategy of ANNs with the energy-saving potential and fast inference capability of SNNs. However, under extremely low-latency conditions, the existing conversion theory suggests that the problem of SNNs' neurons firing more or fewer spikes within each layer, i.e., residual error, leads to a performance gap in the converted SNNs compared with the original ANNs. This severely limits the possibility of the practical application of SNNs on delay-sensitive edge devices. Existing conversion methods addressing this problem usually involve modifying the state of the conversion spiking neurons. However, these methods do not consider their adaptability and compatibility with neuromorphic chips. We propose a new approach based on explicit modeling of residual errors as additive noise. The noise is incorporated into the activation function of the source ANN, effectively reducing the impact of residual error on SNN performance. Our experiments on the CIFAR10/100 and Tiny-ImageNet datasets verify that our approach exceeds the prevailing ANN-SNN conversion methods and directly trained SNNs concerning accuracy and the required time steps. Overall, our method provides new ideas for improving SNN performance under ultralow-latency conditions and is expected to promote practical neuromorphic hardware applications for further development. The code for our NQ framework is available at https://github.com/hzp2022/ANN2SNN_NQ.

Green communications - the ability to obtain a given communication service at minimal energy use - is of growing importance. In addition, future releases (Release-10 and beyond) of the fourth generation (4G) wireless networks are expected to carry an ever increasing number of broadband services. As such, the optimal configuration of 4G networks in the context of minimizing energy usage is very important. In this work we investigate the use of optimal power allocations in 4G networks in the context of co-operative communications. Our system models, which are based on emerging 4G standards, consist of multiple users communicating with a base station through the use of an additional relay station. We show, through analysis and simulations, how optimal power allocation schemes at the relay can lead to significant improvement in throughput. Such increases in throughput, in turn, lead to dramatic decreases in the energy usage of the network. We highlight two main energy savings that can be achieved without impacting any communication service; the opportunity to power-down base stations, and reduced energy consumption at the relay. We show how factor of two savings in energy usage can be expected from the deployment of our techniques.

Development of spiking neural networks (SNNs) controlling mobile robots is one of the modern challenges in computational neuroscience and artificial intelligence. Such networks, being replicas of biological ones, are expected to have a higher computational potential than traditional artificial neural networks (ANNs). The critical problem is in the design of robust learning algorithms aimed at building a “living computer” based on SNNs. Here, we propose a simple SNN equipped with a Hebbian rule in the form of spike-timing-dependent plasticity (STDP). The SNN implements associative learning by exploiting the spatial properties of STDP. We show that a LEGO robot controlled by the SNN can exhibit classical and operant conditioning. Competition of spike-conducting pathways in the SNN plays a fundamental role in establishing associations of neural connections. It replaces the irrelevant associations by new ones in response to a change in stimuli. Thus, the robot gets the ability to relearn when the environment changes. The proposed SNN and the stimulation protocol can be further enhanced and tested in developing neuronal cultures, and also admit the use of memristive devices for hardware implementation.

Inspired by the remarkable properties of the human brain, the fields of machine learning, computational neuroscience and neuromorphic engineering have achieved significant synergistic progress in the last decade. Powerful neural network models rooted in machine learning have been proposed as models for neuroscience and for applications in neuromorphic engineering. However, the aspect of robustness is often neglected in these models. Both biological and engineered substrates show diverse imperfections that deteriorate the performance of computation models or even prohibit their implementation. This thesis describes three projects aiming at implementing robust learning with local plasticity rules in neural networks. First, we demonstrate the advantages of neuromorphic computations in a pilot study on a prototype chip. Thereby, we quantify the speed and energy consumption of the system compared to a software simulation and show how on-chip learning contributes to the robustness of learning. Second, we present an implementation of spike-based Bayesian inference on accelerated neuromorphic hardware. The model copes, via learning, with the disruptive effects of the imperfect substrate and benefits from the acceleration. Finally, we present a robust model of deep reinforcement learning using local learning rules. It shows how backpropagation combined with neuromodulation could be implemented in a biologically plausible framework. The results contribute to the pursuit of robust and powerful learning networks for biological and neuromorphic substrates.

Micro-climatic variations and their impact on domestic energy consumption – Systematic literature review

Unleashing the potential of spiking neural networks for epileptic seizure detection: A comprehensive review

Highly efficient performance-resources trade-off of the biological brain is a motivation for research on neuromorphic computing. Neuromorphic engineers develop event-based spiking neural networks (SNNs) in hardware. Learning in SNNs is a challenging topic of current research. Reinforcement learning (RL) is a particularly promising learning paradigm, important for developing autonomous agents. In this paper, we propose a digital multiplier-less hardware implementation of an SNN with RL capability. The network is able to learn stimulus-response associations in a context-dependent learning task. Validated in a robotic experiment, the proposed model replicates the behavior in animal experiments and the respective computational model. Index Terms-Neuromorphic engineering, spiking neural networks, reinforcement learning, context-dependent task.

Effect of winter feeding systems on farm greenhouse gas emissions

In perimeter building zones with glass façades, controllable fenestration (daylighting/shading) and electric lighting systems are used as comfort delivery systems under dynamic weather conditions, and their operation affects daylight provision, outside view, lighting energy use, as well as overall occupant satisfaction with the visual environment. A well-designed daylighting and lighting control should be able to achieve high level of satisfaction while minimizing lighting energy consumption. Existing daylighting control studies focus on minimizing energy use with general visual comfort constraints, when adaptive and personalized controls are needed in high performance office buildings. Therefore, reliable and efficient models and methods for learning occupants’ personalized visual preference or satisfaction are required, and the development of optimal daylighting controls requires integrated considerations of visual preference/satisfaction and energy use. In this Dissertation, a novel method is presented first for developing personalized visual satisfaction profiles in daylit offices using Bayesian inference. Unlike previous studies based on action data, a set of experiments with human subjects was designed and conducted to collect comparative visual preference data (by changing visual conditions) in private offices. A probit model structure was adopted to connect the comparative preference with a latent satisfaction utility model, assumed in the form of a parametrized Gaussian bell function. The distinct visual satisfaction models were then inferred using Bayesian approach with preference data. The posterior estimations of model parameters, and inferred satisfaction utility functions were investigated and compared, with results reflecting the different overall visual preference characteristics discovered for each person. Second, we present an online visual preference elicitation learning framework for efficiently learning and eliciting occupants’ visual preference profiles and hidden satisfaction utilities. Another set of experiments with human subjects was conducted to implement the proposed learning algorithm in order to validate the feasibility of the method. A combination of Thompson sampling and pure exploration (uncertainty learning) methods was used to balance exploration and exploitation when targeting the near-maximum area of utility during the learning process. Distinctive visual preference profiles of 13 subjects were learned under different weather conditions, demonstrating the feasibility of the learning framework. Entropy of the distribution of the most preferred visual condition is computed for each learned preference profile to quantify the certainty. Learning speed varies with subjects, but using a single variable model (vertical illuminance on the eye), most subjects could be learned to an acceptable certainty level within one day of stable weather, which shows the efficiency of the method (learning outcomes). Finally, a personalized shading control framework is developed to maximize occupant satisfaction while minimizing lighting energy use in daylit offices with roller shades. An integrated lighting-daylighting simulation model is used to predict lighting energy use while it also provides inputs for computing personalized visual preference profiles, previously developed using Bayesian inference from comparative preference data. The satisfaction utility and the predicted lighting energy use are then used to form an optimization framework. We demonstrate the results of: (i) a single objective formulation, where the satisfaction utility is simply used as a constraint to when minimizing lighting energy use and (ii) a multi-objective optimization scheme, where the satisfaction utility and predicted lighting energy use are formulated as parallel objectives. Unlike previous studies, we present a novel way to apply the MOO without assigning arbitrary weights to objectives: allowing occupants to be the final decision makers in real-time balancing between their personalized visual satisfaction and energy use considerations, within dynamic hidden optimal bounds – through a simple interface. In summary, we present the first method to incorporate personalized visual preferences in optimal daylighting control, with energy use considerations, without using generic occupant behavior models or discomfort-based assumptions.

Financial Time Series Forecasting: Comparison of Traditional and Spiking Neural Networks

The success of deep learning methods over the past decade has been partially shrouded in the shadow of adversarial attacks. Even a tiny undetectable deformation can lead to vicious misleading targeted at safety-critical applications. In contrast, the brain is far more robust when performing complex cognitive tasks. Nevertheless, the underlying mechanisms that contribute to the brain’s high reliability remain largely unexplored. At the intersection of neuroscience and artificial intelligence, we show that neuromorphic paradigms offer a promising solution to the dilemma brought by deep learning’s inherent vulnerabilities. Specifically, we exploit the temporal processing capabilities of spiking neural networks (SNNs) to achieve robustness surpassing that of traditional artificial neural networks (ANNs). We demonstrate that prioritizing task-critical information in the encoded sequence and employing early exit decoding to ignore later perturbations significantly enhance SNN robustness. Further improvements in robustness are achieved by accurately capturing temporal dependencies through specialized training algorithms. Additionally, we introduce a fusion encoding strategy to balance SNN generalization on natural data with robustness against adversarial input. Experimental results on the CIFAR-10 dataset show that SNNs trained with these combined methods achieve twice the robustness of ANNs. Overall, our work demonstrates that neuromorphic computing, leveraging the temporal processing capabilities of SNNs, not only provides superior robustness compared to ANNs but also retains the benefit of low energy consumption. These advancements pave the way for developing next-generation, environmentally friendly, and reliable spike-based intelligent systems.

Neuromorphic engineers develop event-based spiking neural networks (SNNs) in hardware. These SNNs closer resemble the dynamics of biological neurons than conventional artificial neural networks and achieve higher efficiency thanks to the event-based, asynchronous nature of the processing. Learning in the hardware SNNs is a more challenging task, however. The conventional supervised learning methods cannot be directly applied to SNNs due to the non-differentiable event-based nature of their activation. For this reason, learning in SNNs is currently an active research topic. Reinforcement learning (RL) is a particularly promising learning method for neuromorphic implementation, especially in the field of autonomous agents' control. An SNN realization of a bio-inspired RL model is in the focus of this work. In particular, in this article, we propose a new digital multiplier-less hardware implementation of an SNN with RL capability. We show how the proposed network can learn stimulus-response associations in a context-dependent task. The task is inspired by biological experiments that study RL in animals. The architecture is described using the standard digital design flow and uses power- and space-efficient cores. The proposed hardware SNN model is compared both to data from animal experiments and to a computational model. We perform a comparison to the behavioral experiments using a robot, to show the learning capability in hardware in a closed sensory-motor loop.