Bio-Inspired Models of Networks, Information, and Computing Systems

As it also happens in nature, technological networks typically evolve in an incremental manner, instead of being optimally designed. This evolutionary process is driven by changes in the underlying parameters and constraints (the “environment”) and it typically aims to minimize the modification cost after each change in the environment. In this paper, we first formulate the incremental network design approach and compare that with the more traditional optimized design approach in which the objective is to minimize the total network cost. We evaluate the cost overhead and evolvability of incremental design under two network expansion models (random and gradual), focusing on the simpler case of “ring” networks. We find that even though incremental design has some cost overhead, that overhead does not increase as the network grows. Also, it is less costly to evolve an existing network than to design it from scratch as long as the network expansion factor is less than a critical value.

In this paper, we address the information routing problem in heterogeneous decision networks with known messages sizes. A decision network is a set of connected nodes in which some nodes are decision makers (DMs) requesting information, others are information providers (sources) or neutral nodes. An information might be relevant to many DMs and can be provided by different sources with different accuracies. The information value for each DM is modelled as a time dependent utility function. The problem is therefore to generate a set of efficient routing plans to satisfy the DMs’ requests. The congestion problem is solved by determining the optimal transmission schedule along the chosen paths. The joint routing-scheduling problem is modelled as a bi-objective optimization problem that maximizes the overall utility and reliability of the generated paths. A multiobjective genetic algorithm (MOGA) is proposed to solve such an NP-hard problem. We show through empirical experiments that the MOGA provides a representative sample of the efficient set. We also develop an upper bound for the first objective, to validate the quality of the generated potentially efficient solutions.

The PPAM is a hardware architecture for a robust, bidirectional and scalable hetero-associative memory. It is fundamentally different from the traditional processing methods which use arithmetic operations and consequently ALUs. In this paper, we present the results of applying the PPAM to a real-world robotics hand-eye coordination task. A comparison is performed with a nearest neighbour technique that was originally used to associate the same dataset. The number of memory load/store operations and the number of ALU operations for the nearest neighbour algorithm is compared with the corresponding PPAM which acheives the same association. It was determined that 29 conflict resolving nodes were required to fully store and recall the entire dataset and the maximum number of memory locations required in any node was 160, with the average and quartiles being much lower.

In this paper, the reliability of a collective robotic system is analysed using two different techniques from the field of reliability engineering. The techniques, Failure Mode and Effect Analysis (FMEA) and Fault Tree Analysis (FTA), are used to analyse and compare two variants of a previously developed ‘autonomous morphogenesis’ controller. The reliability of the controller is discussed and areas where improvements could be made are suggested. The usefulness of FMEA and FTA as aids to the design of fault tolerant collective robotic systems, and the comparative effectiveness of the two approaches, is also discussed.

This paper discusses the notion of “core bio-inspired services” - low-level services providing basic bio-inspired mechanisms, such as evaporation, aggregation or spreading - shared by higher-level services or applications. Design patterns descriptions of self-organising mechanisms, such as gossip, morphogenesis, or foraging, show that these higher-level mechanisms are composed of basic bio-inspired mechanisms (e.g. digital pheromone is composed of spreading, aggregation and evaporation). In order to ease design and implementation of self-organising applications (or high-level services), by supporting reuse of code and algorithms, this paper proposes BIO-CORE, an execution model that provides these low-level services at the heart of any middleware or infrastructure supporting such applications, and provides them as “core” built-in services around which all other services are built.

Memristors are used to compare three gathering techniques in an already-mapped environment where resource locations are known. The All Site model, which apportions gatherers based on the modeled memristance of that path, proves to be good at increasing overall efficiency and decreasing time to fully deplete an environment, however it only works well when the resources are of similar quality. The Leafcutter method, based on Leafcutter ant behaviour, assigns all gatherers first to the best resource, and once depleted, uses the All Site model to spread them out amongst the rest. The Leafcutter model is better at increasing resource influx in the short-term and vastly out-performs the All Site model in a more varied environments. It is demonstrated that memristor based abstractions of gatherer models provide potential methods for both the comparison and implementation of agent controls.

We revisit previous analyses on the computation of the maximum mutual information between a genetic sequence and its mutated versions down the generations, taking into account the protein translation mechanism of the genetic machinery. This amounts to the application of Shannon’s capacity to the study of the transmission of genetic information. Studies on this subject were started by Yockey and then followed by a number of researchers. Here we refine prior analyses employing the Kimura model of base substitution mutations, which is more realistic than the Jukes-Cantor model used by all previous research on this topic. Furthermore we undertake exact computations where prior works just used approximations, and we propose two practical applications of genetic capacity.

The Hawk and Dove game is a well known model from biology for competition over resources between two types of behaviors: aggressive (Hawk) and peaceful (Doves). The game allows to predict whether one of the behaviors will dominate the other or whether we may expect coexistence of both at a long run; in the latter case it allows to predict what fraction of the population will be aggressive and what peaceful. This game is quite relevant to networking, and has been used in the past to predict the outcome of competition between congestion [2] control protocols (both in wireline and in wireless) as well as between power control protocols for wireless communications. In this paper we study new aspects of the game within the framework of flocks of birds, and obtain results that can be useful for network engineering applications as well.

Vehicle routing has traditionally been considered from an optimisation perspective in relation to minimising costs associated with distance travelled and number of vehicles used. Increasingly however, there is a demand to additionally optimise journeys according to the levels of carbon emissions. Incorporating this criterion into an optimisation algorithm necessitates the use of vehicle emission models. Although a number exist, they are often complex to use and customise due to the number of parameters involved. In this paper, we evaluate the use of an Evolutionary Algorithm to calibrate the parameters of a vehicle emissions model against real-world observed emissions data . The calibrated model can then be used with confidence within the fitness function of an optimisation technique on similar data. Initial results obtained suggest that this approach shows promise. The work forms the initial stages of a wider programme of work to investigate the use of nature inspired methods to construct accurate vehicle emissions models for use in green logistics.

Most of the Business Intelligence tools available on the market today have either been developed and industrially operationalised as “one size fits all” solutions or offered with multiple options leaving the business to decide on the best technology to use. We infer that this approach is likely to result in various analysis inaccuracies; hence rendering inappropriate business decisions. Accordingly, evaluating which technologies present more accurate results against a particular business need remains imperative. While using customer data from a large financial services company in South Africa, we analysed the performance of Neural Networks, Artificial Immune Systems and Bayesian Networks in classifying customer buying patterns. We measured the accuracy percentage values for a customer’s propensity to buy policies and also for existing policies lapsing. We observed that such assessments provide great insight in assessing the effectiveness of Business Intelligence enabling technologies. In particular, when applied to a larger data set, various customer patterns can be unearthed which results in adequate customer segmentation and business lead optimisation.

In this paper a biologically-inspired model for recognition of overlapped patterns is proposed. Information processing along the two visual information processing pathways, i.e., the dorsal and the ventral pathway, is modeled as a solution to the problem. We hypothesize that dorsal pathway, in addition to encoding the spatial information, learns the shape representations of the patterns. In our model dorsal pathway uses shape knowledge as a top-down guidance signal to segment the bottom-up saliency map of the overlapped patterns. Segmented map is used to modulate processing in the ventral pathway. Pattern segmentation in the dorsal pathway is implemented as an interactive process. Interaction between the two pathways leads to sequential recognition of the overlapped patterns along the ventral pathway, one after another. Simulation results support the presented hypothesis as well as effectiveness of the model.

Swarm robots provide greater flexibility and robust performance in tasks such as sensing and monitoring of unstructured and unpredictable environments. They need to spread out in these environments maximizing coverage and maintaining network connectivity for efficient operation. Inspired from nature, we design a new coverage and connectivity maintenance algorithm. The algorithm is based on the local rules used by fish while schooling. Each robot is subject to three forces: a) A separation force that pushes it away from its neighbours and increases the size of the swarm. b) A cohesion force that maintains the connectivity of the swarm. c) An alignment force that keeps it aligned to its neighbours and makes relocation faster. Empirical analysis shows that our new algorithm improves coverage and maintains connectivity. Moreover, preliminary results obtained from the basic experiments show that the new swarm-based algorithm outperforms even the most prominent state-of-the-art algorithms, achieving better and faster coverage.

This paper addresses the problem of network monitoring by investigating the performance of three soft computing techniques, the Artificial Neural Network, Bayesian Network and the Artificial Immune System. The techniques were used for achieving situation recognition and monitoring in a large network of Wi-Fi hotspots as part of a highly scalable preemptive monitoring tool for wireless networks. Using a set of data extracted from a live network of Wi-Fi hotspots managed by an ISP, we integrated algorithms into a data collection system to detect anomalous performance and aberrant behavior in the ISP’s network. The results are therefore revealed and discussed in terms of both anomaly performance and aberrant behavior on several test case scenarios.

The energy conservation in chemical sensor networks is crucial as chemical sensors with air sampling to consume significant energy for sensing activity compared to that used for communication unlike other types of sensors, such as optical or acoustic. When considering the threat environment, the chemical tracers dispersed by turbulent motion in the environment display rather complex and even chaotic properties. Hazardous chemical releases are rare events. If all sensors in a wireless chemical sensor network (WCSN) are left in the active state continuously, it will result in significant power consumption. Therefore, dynamic sensor activation is essential for the durability of WCSNs. Dynamic sensor activation for chemical sensor networks using an epidemiology-based sensor activation protocol has been proposed in the literature. In this paper, we investigate the performance of a variant of epidemiology, gossip inspired sensor activation protocol of a WCSN in a chemical tracer field. The simulation framework with gossip protocol is validated against an analytical model. We then perform simulation experiments to evaluate the performance of gossip- based sensor activation protocol on selected performance metrics: number of active sensors and reliability of detection. We show by simulations that by varying the communication radii of sensors, we can achieve better energy conservation while maintaining better performance of a WCSN with a gossip-based activation protocol.

Nanotechnology is currently being applied to vast number of fields to overcome the challenges faced with existing technologies, which can not efficiently scale down to nano level. Communication is one of the important problems to be addressed in nano scale environment, and molecular communication is a candidate to address this problem. Existing research on molecular communication concentrates on application of existing digital communication paradigm. In this paper, we approach the problem from another perspective, and propose an analog communication model in which the data is not quantified. The proposed model enables achieving higher data rates using less energy while keeping the error rate bounded. With this characteristics, the proposed method finds promising application options for specific set of communication requirements.

Molecular communication is the process of transmitting information by modulating the concentration of molecules over time. Molecular communication is suitable for autonomous nanomachines which are limited in size and capability and for interfacing with biological systems which perform functions controlled or influenced by molecules. Some functions may require nanomachines to perform sequential processes. Molecular communication can be used to synchronize multiple nanomachines and to coordinate the timing of the functionality. In this paper, transmitters self-oscillate by releasing a spike of negative autoregulating molecules when concentration of the molecule is below a threshold. When the concentration from a spike disperses and decreases below the threshold, the transmitter releases another spike of molecules. When the environment includes two transmitters, the oscillations of the two transmitters achieve in-phase or anti-phase synchronization depending on the distance between the transmitter and receiver. When there are multiple transmitters arranged in a circle, the oscillations of the transmitters produce in-phase or partially in-phase synchronization. Simulations were performed to characterize the period of oscillation and the phase difference in the oscillations of multiple transmitters.

In the emerging area of nanonetworks, nano to micro-scale devices called nanomachines are deployed to perform various tasks for applications [1][2]. In this short paper, we consider a biosensor network that consists of nano-scale biosensors; i.e., sensors capable of sensing chemical signals (e.g., toxic chemical substances). In the biosensor network, stationary sensors are distributed over a two dimensional space, and expected to capture a chemical signal that appears in the space and that propagates via Brownian motion. We employ two different placement schemes to distribute sensors, and measure the propagation delay that is required to detect a chemical signal. Preliminary simulation results are provided to show the impact of placement schemes as well as the number of sensors on the propagation delay.

In this paper, we show the detrimental effects of Co-Channel Interference (CCI) in Molecular Communications in the context of the Communication via Diffusion (CvD) system. The effects of CCI are evaluated with respect to system performance parameters, probability of hitting to the intended receiver and the channel capacity, while considering additional environmental affects such as the Inter symbol Interference (ISI). Based on our simulation results in a 3D diffusion environment, we conclude that similar to classical wireless communication systems, CCI is an important source that adversely affects the performance in a CvD system. Also, we show that a molecular reuse range concept which is analogous to the frequency reuse range in wireless communications, can be used to cope up and control the severity of CCI where necessary.

In this paper, a guideline is provided for design and optimization of the shape of active transport molecular communication channels. In particular rectangular channels are considered and it is shown that for channels employing a single microtubule as the carrier of the information particles, the smaller the perimeter of the channel, the higher the channel capacity. Furthermore, it is shown that when channels with similar perimeters are considered, square-like channels achieve higher channel capacity for small values of time per channel use, while narrower channels achieve higher information rates for larger values of time per channel use.

Cloud computing is considered the new paradigm in computing that would make computing a utility. Once the cloud computing becomes available widespread, many service providers would market their services at different qualities and prices. When this happens, the customers would be required to select the right service provider who could meet their anticipated quality. A trust management system would identify the quality of service providers and help customers to choose the right provider. Designing a trust management system is a difficult task, as it requires the consideration of several attributes both local and external to the system. In this paper, the authors propose that the Bees Algorithm that was used to solve issues in diverse fields could be successfully adapted to address the trust issue in the cloud computing system. The authors justify their proposition based on the comparative study carried out on cloud computing and the bees environments.

Despite numerous theoretical investigations on Artificial negative selection (ANS), there are still no useful scenarios in which this paradigm would outperform mainstream machine learning, statistical classification methods or other bio-inspired classification approaches. The aim of this paper is to identify main characteristics and requirements of a useful ANS scenario. Our investigations on this question led us to the need to extend the original ANS model proposed by Forrest et al. in [4]. The motivation of our work relies on the observation that biological mechanisms are not isolated mechanisms with a broad application range. They are only suitable for highly specific tasks and they might only be efficient in interaction with the rest of the biological environment.

Computational modeling of biological systems is becoming increasingly important in the endeavors to better understand complex biological behavior. It enables researchers to perform computerized simulations using a systems biology approach, in order to understand the underlying mechanisms of certain biological phenomena. It provides an opportunity to perform experiments that are otherwise impractical or infeasible in vivo/vitro experiments. In our approach we propose to model and simulate the pathogenesis ofMycobacterium marinum using Petri Net formalism based on data obtained from analysis of microscope images and to provide a three dimensional visualization of the whole infection process and granuloma formation. Image analysis will provide an accurate estimation of the infection in a structured database which will be used for the construction of the Petri Net model. The results of the simulation and analysis of the infection behavior will be visualized in 3D.

This paper explores approaches to Adaptive Information Filtering (AIF) in the context of changing user interests. Based on the existing artificial immune system for email classification (AISEC), we demonstrate an effective extension to classification based on the body of emails. Widening this to the problem of AIF on dynamic web content, we propose to explore dynamic clonal selection algorithms (DCSAs) that include dynamically changing thresholds.

This paper describes improvements in both the efficiency and scalability of a computer simulation originally developed to illustrate idiotypic network theory in the immune system. The original synchronous model was streamlined using software engineering principles to improve efficiency before being adapted to function as a multi threaded application using asynchronous communication that better mimics the decentralised nature of the the biological system that it demonstrates.

Describing, understanding, and modeling the behavior of systems built upon self-organizing principles (such as many bio-inspired systems) is key to engineering self-organizing systems that can solve problems in real computing environments. Capturing the properties of the micro-macro linkage that connects local behaviors of system components to global emergent properties of the system as a whole is particularly important. Different kinds of models have been proposed, each focusing on a different aspect of the problem: descriptive models provide notations that support the design activity and the application of self-organzing principles; validation models allow formal examination of dynamic properties; and analytic models provide techniques for mathematical exploration of abstracted collective behaviors. Our goal is to identify and select the best tools available from these families, extend them where needed, and tie them together to support the creation and analysis of self-organized autonomic computing systems in an integrated way.