A diffusion molecular channel is a channel with memory, as molecules released into the medium hit the receptors after a random delay. Coding over the diffusion channel is performed by choosing the type, intensity, or the released time of molecules diffused in the environment over time. To avoid intersymbol interference (ISI), molecules of the same type should be released at time instances that are sufficiently far apart. This ensures that molecules of a previous transmission are faded in the environment, before molecules of the same type are reused for signaling. In this paper, we consider ISI-free timeslotted modulation schemes. The maximum reliable transmission rate for these modulations is given by the constrained coding capacity of the graph that represents the permissible transmission sequences.However, achieving the constrained coding capacity requires long blocklengths and delays at the decoder, making it impractical for simple nanomachines. The main contribution of this paper is to consider modulations with small delay (short blocklength) and show that they get very close to constrained coding capacity. June 13, 2018 DRAFT arXiv:1508.02495v1 [cs.IT] 11 Aug 2015 2 I. INTRODUCTION Molecular communications (MC) via diffusion is one of the most promising approaches for communications among nanonetworks [1], [2]. In this approach, information is conveyed to a nano-receiver by nano-transmitter's choice of the concentration, type, or the release time of the molecules diffused into the medium. The motion of the released molecules are described by a Brownian motion process, that can be with or without drift. A molecule released from a transmitter can follow different trajectories before reaching a receiver. Thus, diffusion-based communication suffers from intersymbol interference (ISI) due to molecules from previous transmission that follow longer trajectories before hitting the receiver. The effect of ISI in diffusion based MC has been studied extensively in the literature [3]-[6], with the general conclusion that ISI reduces the performance in a communication setup consisting of one transmitter and one receiver. The authors in [7], [8] have proposed two modulation schemes, Concentration Shift Keying (CSK) and Molecular Shift Keying (MOSK), where both suffer from the ISI caused by molecules from previous transmissions. Authors in [9] have proposed the Molecular Concentration Shift Keying (MCSK) modulation scheme, where the main idea is to use the distinct molecule types for consecutive time slots at the transmitter, thus effectively suppressing the ISI. In [10] the authors take advantage of a limited memory in the transmitter to propose an adaptive transmission rate scheme that reduces ISI. An on-off MOSK modulation scheme has been proposed in [11] which increases the transmission rate compared to MOSK but still suffers from the ISI. In [12], a modulation, called run-length hybrid aware, has been presented, which considers the runs (the same value occurs in several consecutive bits) for encoding: the run-value is...

No abstract

In this paper, a simple memory limited transmitter for molecular communication is proposed, in which information is encoded in the diffusion rate of the molecules. Taking advantage of memory, the proposed transmitter reduces the ISI problem by properly adjusting its diffusion rate. The error probability of the proposed scheme is derived and the result is compared with the lower bound on error probability of the optimum transmitter. It is shown that the performance of introduced transmitter is near optimal (under certain simplifications). Simplicity is the key feature of the presented communication system: the transmitter follows a simple rule, the receiver is a simple threshold decoder and only one type of molecule is used to convey the information. I. INTRODUCTIONNew applications such as smart drug delivery and health monitoring give rise to the importance of molecular communication, a new paradigm for communication between nanomachines over a short (nanoscale or microscale) range. In molecular communication, information is carried by molecules, rather than electrons or electromagnetic waves [1], [2]. Several types of molecular communication have been considered, among them, diffusion based communication, which corresponds to traditional wireless communication [3], is of great interest, since it does not require any prior communication link infrastructure. In diffusion based communication, the transmitter nanomachine releases information molecules in the environment. These released molecules diffuse randomly until they hit the receiver nanomachine.[4], [5]. Due to the random nature of molecular propagation, diffusion based communication suffers from inter symbol interference (ISI). Several solutions have been proposed to mitigate ISI (e.g. see [6]-[9]). In [10], a new modulation technique, named Molecular Concentration Shift Keying (MCSK), is suggested. Exploiting two types of molecules, while MCSK eliminates the interference from the last transmitted symbol and reduces the error probability, it suffers from interference due to earlier transmissions. A solution based on adding intelligence to receiver is suggested in [11]. where the receiver stores the last decoded bits in memory to make an estimation of current interference level, and uses this estimation to adjust the threshold for decoding the current bit. In [12], a linear and time invariant model is presented and the optimal receiver is derived, under this model. However, this receiver is too complex to be implemented in practice. In [13], the authors considered a deterministic noiseless diffusion channel with memory, and proposed using different symbol durations to deal with ISI by taking into account the channel binary concentration state. They then computed the channel capacity by adapting the Shannon telegraph channel method. In this paper, we propose a simple transmitter which significantly reduces ISI by adaptively adjusting transmission rates to stabilize the rate of molecules at the receiver, enabling the use of a simple fixed threshold receiver. To...

No abstract

No abstract