Spike-timing-dependent plasticity in balanced random networks

The balanced random network model attracts considerable interest because it explains the irregular spiking activity at low rates and large membrane potential fluctuations exhibited by cortical neurons in vivo. In this article, we investigate to what extent this model is also compatible with the experimentally observed phenomenon of spike-timing-dependent plasticity (STDP). Confronted with the plethora of theoretical models for STDP available, we reexamine the experimental data. On this basis, we propose a novel STDP update rule, with a multiplicative dependence on the synaptic weight for depression, and a power law dependence for potentiation. We show that this rule, when implemented in large, balanced networks of realistic connectivity and sparseness, is compatible with the asynchronous irregular activity regime. The resultant equilibrium weight distribution is unimodal with fluctuating individual weight trajectories and does not exhibit development of structure. We investigate the robustness of our results with respect to the relative strength of depression. We introduce synchronous stimulation to a group of neurons and demonstrate that the decoupling of this group from the rest of the network is so severe that it cannot effectively control the spiking of other neurons, even those with the highest convergence from this group.     

Response time variability

Response time variability is a new optimization problem with a broad range of applications and a distinctive number of theoretic flavour. The problem occurs whenever events, jobs, clients or products need to be sequenced so as to minimize the variability of time for which they wait for the next turn in obtaining the resources necessary for their advance. The problem has numerous real-life applications. We study its computational complexity, present efficiency, polynomial time algorithms for some cases, and the NP-hardness proof for a general problem. We propose a position exchange heuristic and apply it to improve the total response time variability of an initial sequence. The latter is the optimum bottleneck sequence, Webster or Jefferson sequence of the apportionment, or a random sequence. We report on computational experiments with the heuristic.     

Computing and stability in cortical networks

Cortical neurons are predominantly excitatory and highly interconnected. In spite of this, the cortex is remarkably stable: normal brains do not exhibit the kind of runaway excitation one might expect of such a system. How does the cortex maintain stability in the face of this massive excitatory feedback? More importantly, how does it do so during computations, which necessarily involve elevated firing rates? Here we address these questions in the context of attractor networks-networks that exhibit multiple stable states, or memories. We find that such networks can be stabilized at the relatively low firing rates observed in vivo if two conditions are met: (1) the background state, where all neurons are firing at low rates, is inhibition dominated, and (2) the fraction of neurons involved in a memory is above some threshold, so that there is sufficient coupling between the memory neurons and the background. This allows "dynamical stabilization" of the attractors, meaning feedback from the pool of background neurons stabilizes what would otherwise be an unstable state. We suggest that dynamical stabilization may be a strategy used for a broad range of computations, not just those involving attractors.     

Correlations and population dynamics in cortical networks

The function of cortical networks depends on the collective interplay between neurons and neuronal populations, which is reflected in the correlation of signals that can be recorded at different levels. To correctly interpret these observations it is important to understand the origin of neuronal correlations. Here we study how cells in large recurrent networks of excitatory and inhibitory neurons interact and how the associated correlations affect stationary states of idle network activity. We demonstrate that the structure of the connectivity matrix of such networks induces considerable correlations between synaptic currents as well as between subthreshold membrane potentials, provided Dale's principle is respected. If, in contrast, synaptic weights are randomly distributed, input correlations can vanish, even for densely connected networks. Although correlations are strongly attenuated when proceeding from membrane potentials to action potentials (spikes), the resulting weak correlations in the spike output can cause substantial fluctuations in the population activity, even in highly diluted networks. We show that simple mean-field models that take the structure of the coupling matrix into account can adequately describe the power spectra of the population activity. The consequences of Dale's principle on correlations and rate fluctuations are discussed in the light of recent experimental findings.     

Exact and approximate balanced data gathering in energy-constrained sensor networks

We consider the problem of gathering data from a wireless multi-hop network of energy-constrained sensor nodes to a common base station. Specifically, we aim to balance the total amount of data received from the sensor network during its lifetime against a requirement of sufficient coverage for all the sensor locations surveyed. Our main contribution lies in formulating this balanced data gathering task, studying the effects of balancing, and proposing an approximation algorithm for the problem. Based on an LP network flow formulation, we present experimental results on both optimal and approximate data routing designs, in open transmission ranges and with impenetrable obstacles between the nodes.     

Approximation schemes for load balanced clustering in wireless sensor networks

Clustering sensor nodes is an efficient technique to improve scalability and life time of a wireless sensor network (WSN). However, in a cluster based WSN, the leaders (cluster heads) consume more energy due to some extra load for various activities such as data collection, data aggregation, and communication of the aggregated data to the base station. Therefore, balancing the load of the cluster heads is a crucial issue for the long run operation of the WSNs. In this paper, we first present a load balanced clustering scheme for wireless sensor networks. We show that the algorithm runs in O(nlogn) time for n sensor nodes. We prove that the algorithm is optimal for the case in which the sensor nodes have equal load. We also show that it is a polynomial time 2-approximation algorithm for the general case, i.e., when the sensor nodes have variable load. We finally improve this algorithm and propose a 1.5-approximation algorithm for the general case. The experimental results show the efficiency of the proposed algorithm in terms of the load balancing of the cluster heads, execution time, and the network life.     

The high-conductance state of cortical networks

We studied the dynamics of large networks of spiking neurons with conductance-based (nonlinear) synapses and compared them to networks with current-based (linear) synapses. For systems with sparse and inhibition-dominated recurrent connectivity, weak external inputs induced asynchronous irregular firing at low rates. Membrane potentials fluctuated a few millivolts below threshold, and membrane conductances were increased by a factor 2 to 5 with respect to the resting state. This combination of parameters characterizes the ongoing spiking activity typically recorded in the cortex in vivo. Many aspects of the asynchronous irregular state in conductance-based networks could be sufficiently well characterized with a simple numerical mean field approach. In particular, it correctly predicted an intriguing property of conductance-based networks that does not appear to be shared by current-based models: they exhibit states of low-rate asynchronous irregular activity that persist for some period of time even in the absence of external inputs and without cortical pacemakers. Simulations of larger networks (up to 350,000 neurons) demonstrated that the survival time of self-sustained activity increases exponentially with network size.     

Swarm optimisation algorithms applied to large balanced communication networks

In the last years, several combinatorial optimisation problems have arisen in the computer communications networking field. In many cases, for solving these problems it is necessary the use of meta-heuristics. An important problem in communication networks is the Terminal Assignment Problem (TAP). Our goal is to minimise the link cost of large balanced communication networks. TAP is a NP-Hard problem. The intractability of this problem is the motivation for the pursuits of Swarm Intelligence (SI) algorithms that produce approximate, rather than exact, solutions. This paper makes a comparison among the effectiveness of three SI algorithms: Ant Colony Optimisation, Discrete Particle Swarm Optimisation and Artificial Bee Colony. We also compare the SI algorithms with several algorithms from literature. Simulation results verify the effectiveness of the proposed algorithms. The results show that SI algorithms provide good solutions in a better running time.     

On the differential and linear efficiency of balanced Feistel networks

Balanced Feistel networks (BFN) have been widely used for constructing efficient block ciphers. They are known to provide high efficiency with respect to differential and linear cryptanalysis, when instantiated with SL-type round functions (BFN-SL). This work suggests that BFNs attain higher efficiency when the round function is defined as a composition of two substitution layers connected by a linear diffusion layer (SLS-type round function). The resulting structure is called BFN-SLS.Tight upper bounds on the differential and linear trail probabilities are proven for such constructions. When compared to BFN-SL with single-round diffusion, BFN-SLS exhibits an increase by almost 1/3 in the proportion of active S-boxes. When compared to BFN-SL with multiple-round diffusion, BFN-SLS provides the same proportion of active S-boxes, requiring, however, twice less linear operations and a single diffusion matrix for all rounds.It is argued that the cost of linear operations cannot be ignored when dealing with efficiency. Different BFNs are compared under consideration of the relative complexity of linear and nonlinear finite field operations. As a result, since BFN-SLS minimizes the number of necessary linear operations, its efficiency is higher than that of the known BFN-SL constructions.     

无线网络负载均衡的干扰感知路由度量

无线信道干扰和负载分布不均匀严重影响无线网络的网络吞吐量、端到端延时等。在已有的路由度量的基础上,充分继承其通过邻居节点负载描述干扰强度的优势,进一步分析节点负载的影响,提出负载均衡的干扰感知路由度量,将干扰邻居节点的数量、负载和距离综合作用结果作为流间干扰强度,使用节点处的平均队列长度捕捉节点负载,并改进期望传输时间消除链路的不对称性,实现负载均衡和干扰感知,避开热点区域。同时将LBIA合并入路由协议。仿真结果表明:该路由度量可以有效地实现网络负载均衡,提升网络整体性能。     

Simulating cortical networks on heterogeneous multi-GPU systems

Recent advances in neuroscientific understanding have highlighted the highly parallel computation power of the mammalian neocortex. In this paper we describe a GPGPU-accelerated implementation of an intelligent learning model inspired by the structural and functional properties of the neocortex. Furthermore, we consider two inefficiencies inherent to our initial implementation and propose software optimizations to mitigate such problems. Analysis of our application’s behavior and performance provides important insights into the GPGPU architecture, including the number of cores, the memory system, atomic operations, and the global thread scheduler. Additionally, we create a runtime profiling tool for the cortical network that proportionally distributes work across the host CPU as well as multiple GPGPUs available to the system. Using the profiling tool with these optimizations on Nvidia’s CUDA framework, we achieve up to 60× speedup over a single-threaded CPU implementation of the model.     

Rate models for conductance-based cortical neuronal networks

Population rate models provide powerful tools for investigating the principles that underlie the cooperative function of large neuronal systems. However, biophysical interpretations of these models have been ambiguous. Hence, their applicability to real neuronal systems and their experimental validation have been severely limited. In this work, we show that conductance-based models of large cortical neuronal networks can be described by simplified rate models, provided that the network state does not possess a high degree of synchrony. We first derive a precise mapping between the parameters of the rate equations and those of the conductance-based network models for time-independent inputs. This mapping is based on the assumption that the effect of increasing the cell's input conductance on its f-I curve is mainly subtractive. This assumption is confirmed by a single compartment Hodgkin-Huxley type model with a transient potassium A-current. This approach is applied to the study of a network model of a hypercolumn in primary visual cortex. We also explore extensions of the rate model to the dynamic domain by studying the firing-rate response of our conductance-based neuron to time-dependent noisy inputs. We show that the dynamics of this response can be approximated by a time-dependent second-order differential equation. This phenomenological single-cell rate model is used to calculate the response of a conductance-based network to time-dependent inputs.     

The complexity of latching transitions in large scale cortical networks

We study latching dynamics, i.e. the ability of a network to hop spontaneously from one discrete attractor state to another, which has been proposed as a model of an infinitely recursive process in large scale cortical networks, perhaps associated with higher cortical functions, such as language. We show that latching dynamics can span the range from deterministic to random under the control of a threshold parameter U. In particular, the interesting intermediate case is characterized by an asymmetric and complex set of transitions. We also indicate how finite latching sequences can become infinite, depending on the properties of the transition probability matrix and of its eigenvalues.     

Scalable balanced training of conditional generative adversarial neural networks on image data

The Journal of Supercomputing - We propose a distributed approach to train deep convolutional generative adversarial neural network (DC-CGANs) models. Our method reduces the imbalance between...     

Distributed hierarchical search for balanced energy consumption routing spanning trees in wireless sensor networks

We provide a new heuristic method approach to search for degree-balanced and small weight routing spanning trees in a network. The method is a modification of Kruskal’s minimum spanning tree search algorithm and is based on a distributed search by hierarchical clusters. It provides spanning trees with a lower maximum weighted degree, a bigger diameter, and can be used for balanced energy consumption routing in wireless sensor networks (WSN’s). The method can be naturally implemented in parallel or as a simple locally distributed algorithm. Simulations for a realistic case scenario WSN are done based on the transmission energy matrix. The simulation results show that the proposed approach can extend the functional lifetime of a WSN in terms of sensor transmission energy by 3–4 times. We also show that the results can be further improved by using a preliminary clustering of the input network.     

Rapid temporal modulation of synchrony by competition in cortical interneuron networks

The synchrony of neurons in extrastriate visual cortex is modulated by selective attention even when there are only small changes in firing rate (Fries, Reynolds, Rorie, & Desimone, 2001). We used Hodgkin-Huxley type models of cortical neurons to investigate the mechanism by which the degree of synchrony can be modulated independently of changes in firing rates. The synchrony of local networks of model cortical interneurons interacting through GABA(A) synapses was modulated on a fast timescale by selectively activating a fraction of the interneurons. The activated interneurons became rapidly synchronized and suppressed the activity of the other neurons in the network but only if the network was in a restricted range of balanced synaptic background activity. During stronger background activity, the network did not synchronize, and for weaker background activity, the network synchronized but did not return to an asynchronous state after synchronizing. The inhibitory output of the network blocked the activity of pyramidal neurons during asynchronous network activity, and during synchronous network activity, it enhanced the impact of the stimulus-related activity of pyramidal cells on receiving cortical areas (Salinas & Sejnowski, 2001). Synchrony by competition provides a mechanism for controlling synchrony with minor alterations in rate, which could be useful for information processing. Because traditional methods such as cross-correlation and the spike field coherence require several hundred milliseconds of recordings and cannot measure rapid changes in the degree of synchrony, we introduced a new method to detect rapid changes in the degree of coincidence and precision of spike timing.     

Accurate prediction of AD patients using cortical thickness networks

It is widely believed that human brain is a complicated network and many neurological disorders such as Alzheimer’s disease (AD) are related to abnormal changes of the brain network architecture. In this work, we present a kernel-based method to establish a network for each subject using mean cortical thickness, which we refer to hereafter as the individual’s network. We construct individual networks for 83 subjects, including AD patients and normal controls (NC), which are taken from the Open Access Series of Imaging Studies database. The network edge features are used to make prediction of AD/NC through the sophisticated machine learning technology. As the number of edge features is much more than that of samples, feature selection is applied to avoid the adverse impact of high-dimensional data on the performance of classifier. We use a hybrid feature selection that combines filter and wrapper methods, and compare the performance of six different combinations of them. Finally, support vector machines are trained using the selected features. To obtain an unbiased evaluation of our method, we use a nested cross validation framework to choose the optimal hyper-parameters of classifier and evaluate the generalization of the method. We report the best accuracy of 90.4 % using the proposed method in the leave-one-out analysis, outperforming that using the raw cortical thickness data by more than 10 %.     

Dynamic genetic algorithms for the dynamic load balanced clustering problem in mobile ad hoc networks

Clustering can help aggregate the topology information and reduce the size of routing tables in a mobile ad hoc network (MANET). To achieve fairness and uniform energy consumption, each clusterhead should ideally support the same number of clustermembers. However, a MANET is a dynamic and complex system and its one important characteristic is the topology dynamics, that is, the network topology changes over time due to the factors such as energy conservation and node movement. Therefore, in a MANET, an effective clustering algorithm should efficiently adapt to each topology change and produce the new load balanced clusterhead set quickly. The maintenance of the cluster structure should aim to keep it as stable as possible to reduce overhead. To meet this requirement, the new solution should keep as many good parts in the previous solution as possible. In this paper, we first formulate the dynamic load balanced clustering problem (DLBCP) into a dynamic optimization problem. Then, we propose to use a series of dynamic genetic algorithms (GAs) to solve the DLBCP in MANETs. In these dynamic GAs, each individual represents a feasible clustering structure and its fitness is evaluated based on the load balance metric. Various dynamics handling techniques are introduced to help the population to deal with the topology changes and produce closely related solutions in good quality. The experimental results show that these GAs can work well for the DLBCP and outperform traditional GAs that do not consider dynamic network optimization requirements.