A Cooperative Machine Learning Approach for Pedestrian Navigation in Indoor IoT.

This paper presents a system based on pedestrian dead reckoning (PDR) for localization of networked mobile users, which relies only on sensors embedded in the devices and device- to-device connectivity. The user trajectory is reconstructed by measuring step by step the user displacements. Though step length can be estimated rather accurately, heading evaluation is extremely problematic in indoor environments. Magnetometer is typically used, however measurements are strongly perturbed. To improve the location accuracy, this paper proposes a novel cooperative system to estimate the direction of motion based on a machine learning approach for perturbation detection and filtering, combined with a consensus algorithm for performance augmentation by cooperative data fusion at multiple devices. A first algorithm filters out perturbed magnetometer measurements based on a-priori information on the Earth's magnetic field. A second algorithm aggregates groups of users walking in the same direction, while a third one combines the measurements of the aggregated users in a distributed way to extract a more accurate heading estimate. To the best of our knowledge, this is the first approach that combines machine learning with consensus algorithms for cooperative PDR. Compared to other methods in the literature, the method has the advantage of being infrastructure-free, fully distributed and robust to sensor failures thanks to the pre-filtering of perturbed measurements. Extensive indoor experiments show that the heading error is highly reduced by the proposed approach thus leading to noticeable enhancements in localization performance.

Using transcription-based detectors to emulate the behavior of sequential probability ratio-based concentration detectors.

The sequential probability ratio test (SPRT) from statistics is known to have the least mean decision time compared to other sequential or fixed-time tests for given error rates. In some circumstances, cells need to make decisions accurately and quickly, therefore it has been suggested that the SPRT may be used to understand the speed-accuracy tradeoff in cellular decision-making. It is generally thought that in order for cells to make use of the SPRT, it is necessary to find biochemical circuits that can compute the log-likelihood ratio needed for the SPRT. However, this paper takes a different approach. We recognize that the high-level behavior of the SPRT is defined by its positive detection or hit rate, and the computation of the log-likelihood ratio is just one way to realize this behavior. In this paper, we will present a method in which a transcription-based detector is used to emulate the hit rate of the SPRT without computing the exact log-likelihood ratio. We consider the problem of using a promoter with multiple binding sites to accurately and quickly detect whether the concentration of a transcription factor is above a target level. We show that it is possible to find binding and unbinding rates of the transcription factor to the promoter's binding sites so that the probability that the amount of mRNA produced will be higher than a threshold is approximately equal to the hit rate of the SPRT detector. Moreover, we show that the average time that this transcription-based detector needs to make a positive detection is less than or equal to that of the SPRT for a wide range of concentrations. We remark that the last statement does not contradict Wald's optimality result because our transcription-based detector uses an open-ended test.

Molecular Communications With Molecular Circuit-Based Transmitters and Receivers.

The performance of a communication link can be improved by maximizing the mutual information between the input and output signals. This paper considers this maximization problem in a molecular communication link where both the transmitter and the receiver are molecular circuit. This general optimization is hard to solve. We simplify the problem by limiting to reactions with linear reaction rates and molecular circuits with a limited number of species. We derive an expression of mutual information and use it for numerical maximization. We show that our parameterized transmitter circuit is able to give mutual information that is close to upper bound obtained in our earlier work.

Detection of persistent signals and its relation to coherent feed-forward loops.

Many studies have shown that cells use the temporal dynamics of signalling molecules to encode information. One particular class of temporal dynamics is persistent and transient signals, i.e. signals of long and short duration, respectively. It has been shown that the coherent type-1 feed-forward loop with an AND logic at the output (or C1-FFL for short) can be used to discriminate a persistent input signal from a transient one. This has been done by modelling the C1-FFL, and then using the model to show that persistent and transient input signals give, respectively, a non-zero and zero output. The aim of this paper is to make a connection between the statistical detection of persistent signals and the C1-FFL. We begin by first formulating a statistical detection problem of distinguishing persistent signals from transient ones. The solution of the detection problem is to compute the log-likelihood ratio of observing a persistent signal to a transient signal. We show that, if this log-likelihood ratio is positive, which happens when the signal is likely to be persistent, then it can be approximately computed by a C1-FFL. Although the capability of C1-FFL to discriminate persistent signals is known, this paper adds an information processing interpretation on how a C1-FFL works as a detector of persistent signals.

Chemical reaction networks for computing logarithm.

Living cells constantly process information from their living environment. It has recently been shown that a number of cell signaling mechanisms (e.g. G protein-coupled receptor and epidermal growth factor) can be interpreted as computing the logarithm of the ligand concentration. This suggests that logarithm is a fundamental computation primitive in cells. There is also an increasing interest in the synthetic biology community to implement analog computation and computing the logarithm is one such example. The aim of this article is to study how the computation of logarithm can be realized using chemical reaction networks (CRNs). CRNs cannot compute logarithm exactly. A standard method is to use power series or rational function approximation to compute logarithm approximately. Although CRNs can realize these polynomial or rational function computations in a straightforward manner, the issue is that in order to be able to compute logarithm accurately over a large input range, it is necessary to use high-order approximation that results in CRNs with a large number of reactions. This article proposes a novel method to compute logarithm accurately in CRNs while keeping the number of reactions in CRNs low. The proposed method can create CRNs that can compute logarithm to different levels of accuracy by adjusting two design parameters. In this article, we present the chemical reactions required to realize the CRNs for computing logarithm. The key contribution of this article is a novel method to create CRNs that can compute logarithm accurately over a wide input range using only a small number of chemical reactions.

Improving the Capacity of Molecular Communication Using Enzymatic Reaction Cycles.

This paper considers the capacity of a diffusion-based molecular communication link assuming the receiver uses chemical reactions. The key contribution is we show that enzymatic reaction cycles, which is a class of chemical reactions commonly found in cells consisting of a forward and a backward enzymatic reaction, can improve the capacity of the communication link. The technical difficulty in analyzing enzymatic reaction cycles is that their reaction rates are nonlinear. We deal with this by assuming that the amount of certain chemicals in the enzymatic reaction cycle is large. In order to simplify the problem further, we use singular perturbation to study a particular operating regime of the enzymatic reaction cycles. This allows us to derive a closed-form expression of the channel gain. This expression suggests that we can improve the channel gain by increasing the total amount of substrate in the enzymatic reaction cycle. By using numerical calculations, we show that the effect of the enzymatic reaction cycle is to increase the channel gain and to reduce the noise, which results in a better signal-to-noise ratio and in turn a higher communication capacity. Furthermore, we show that we can increase the capacity by increasing the total amount of substrate in the enzymatic reaction cycle.

Extended master equation models for molecular communication networks.

We consider molecular communication networks consisting of transmitters and receivers distributed in a fluidic medium. In such networks, a transmitter sends one or more signaling molecules, which are diffused over the medium, to the receiver to realize the communication. In order to be able to engineer synthetic molecular communication networks, mathematical models for these networks are required. This paper proposes a new stochastic model for molecular communication networks called reaction-diffusion master equation with exogenous input (RDMEX). The key idea behind RDMEX is to model the transmitters as time series of signaling molecule counts, while diffusion in the medium and chemical reactions at the receivers are modeled as Markov processes using master equation. An advantage of RDMEX is that it can readily be used to model molecular communication networks with multiple transmitters and receivers. For the case where the reaction kinetics at the receivers is linear, we show how RDMEX can be used to determine the mean and covariance of the receiver output signals, and derive closed-form expressions for the mean receiver output signal of the RDMEX model. These closed-form expressions reveal that the output signal of a receiver can be affected by the presence of other receivers. Numerical examples are provided to demonstrate the properties of the model.