Efficient algorithms and performance bounds for clock synchronization in Wireless Sensor Networks

Wireless Sensor Networks (WSNs) have extensively been worked upon due to their ability to monitor various physical phenomenon. Time synchronization is the most fundamental and most important challenge faced by WSNs. Time Synchronization in WSN can be achieved either by gaining consensus amongst all the nodes or by a reference node which broadcasts a single unit of time to the entire network. The paper describes a novel technique to achieve fully distributed cluster clock synchronization, which synchronizes all the nodes in the network to a common consensus value. The algorithm employs clustering and confidence weighted running average (CWA) method for offset and skew compensations. The efficiency of protocol is validated with simulation results presented in the paper.

It is important and challenging to achieve secure time synchronization in wireless sensor networks, which may be deployed in a hostile environment under various malicious attacks. The recently developed average-consensus-based time synchronization protocol (ATS) is a promising alternative as it does not depend on any reference node or network topology, which makes it robust to different kinds of attacks, e.g., denial-of-service, node destruction, etc. However, the in-network information fusion nature of average consensus makes the ATS vulnerable to the well-known message manipulation attacks. In this paper, we focus on how to defend the ATS protocol in wireless sensor networks under message manipulation attacks. We first investigate the impact of message manipulation attacks over ATS, and derive a necessary condition for ATS to converge. Then based on the obtained insights, we propose a novel adjusting parameter checking mechanism which exploits the two-hop neighboring information to dynamically constrain the attackers. We further incorporate all checking processes into the traditional ATS protocol to form a secure average-consensus-based time synchronization protocol (SATS). We prove that SATS guarantees the network time synchronization with an exponentially converging speed.

Wireless sensor networks require time synchronization, which is the coordination of events or actions to make a system operate in unison. In this work, real experiments and a theoretical analysis of the behavior of the clock sources, most used in wireless sensor networks, have been carried out. The experiments have been performed on two real platforms from two different manufacturers in real environments with sudden changes in temperature. Complementary metal-oxide-semiconductor oscillators have a low accuracy, bigger than 500 ppm, and a high dependency with temperature. External crystal oscillators have good accuracy, around 20 ppm, and are stable with temperature. Temperature-compensated crystal oscillators are very accurate, around 5 ppm, and the temperature has no influence in their drift. The use of phase-locked loop circuits minimizes the impact of temperature and stabilizes oscillators. We highlight and demonstrate the importance of the early stages of design, especially the selection of the clock source, because that decision has a great impact on the performance of the time synchronization in wireless sensor networks.

Motivated by the importance of the clock synchronization in wireless sensor networks (WSNs), this paper proposes a new research approach and model approach, which quantitatively analyzes clock synchronization from the perspective of modern control theory. Two kinds of control strategies are used as examples to analyze the effect of the control strategy on clock synchronization from different perspectives, namely, the single‐step optimal control and the LQG global optimal control. The proposed method establishes a state space model for clock relationship, thus making dimension extension and parameter identification easier, and is robust to changes under the condition of node failures and new nodes. And through the design of different control strategies and performance index functions, the method can satisfy various requirements of the synchronization precision, convergence speed, energy consumption and the computational complexity, and so on. Finally, the simulations show that the synchronization accuracy of the proposed method is higher than that of the existing protocol, and the former convergence speed of the synchronization error is faster.

Study of consensus-based time synchronization in wireless sensor networks

Time synchronization, aiming to provide a common timescale among distributed sensor nodes, is a key enabling technology for many applications, such as collaborative condition monitoring and localization detection. Due to the complexity of time synchronization in wireless sensor networks, the Discrete Event Simulator is recommended to adopt resulting from the feature that the behavior of a complex is represented as an order sequence of well-defined event in time. In this paper, the PCO clock is firstly implemented into the open-source software simulator OMNeT++ by using desynchronization mechanism under a realistic scenario, and it can also be simulated at an adjustable and higher resolution. The developed relay node, can either be a Full Function Device or a Reduced Function Device, enables the high scalability for multi-node and multi-hop simulation. In addition, the shared code of this project on the GitHub directly benefits the researchers and engineers in communication.

Synchronization of tiny sensor nodes forming Wireless Sensor Networks (WSNs) is a challenging problem due to frequent topological changes, node failures and power, memory and computation constraints. These difficulties promote a self-organizing solution for the problem of time synchronization in WSNs to be quite desirable. Current self-organizing time synchronization protocols in WSNs have drawbacks: They either provide synchronicity but not a common notion of time for the nodes in the network or they demand keeping track of the time information of the neighboring nodes. The latter drawback becomes quite crucial especially on dense WSNs, which makes available self-organizing time synchronization protocols impractical. This paper provides a novel self-organizing time synchronization protocol for WSNs which does not require keeping track of the neighboring nodes. The main component of our protocol is a computationally light "adaptive-value tracking" algorithm which synchronizes the rate of each sensor node to that of all its neighboring nodes through successive feedbacks. We show by simulations that the proposed protocol achieves a tight synchronization which results in the desired property of global time notion after a finite amount of time on dense networks. Although there is no formal proof yet of the protocol's convergence, we anticipate this protocol could be used as a practical time synchronization approach.

ABSTRACTThe clock in embedded systems usually is driven by a crystal oscillator and implemented via a counter register, such a crystal clock is non-identical and drifting due to the manufacturing tolerance and variation of working conditions. Thus, a common time among distributed wireless sensor nodes, also referred to as Time Synchronization, is required for many time-sensitive wireless applications, such as collaborative condition monitoring, coordinated control and localization. Inspired by fireflies’ behaviour, the Pulse-Coupled Oscillators (PCO) has been proposed for synchronization in complex networks. Since the concurrent transmission of PCO’s Pulses is impossible in Wireless Sensor Networks (WSNs), the desynchronization mechanism is adopted to ensure the implementation of PCO in WSNs. Moreover, due to the uncertainties in radio channels and the complexities of communication protocols and packet-exchange behaviours in wireless networks, it is challenging to have a closed-form solution to the performance of PCO synchronization in WSNs. The realistic software simulation, in particular, the discrete event simulator has been a powerful tool to exam the performance of communication protocols in various scenarios, since an order sequence of well-defined event in time is to represent the behaviour of a complex system. This paper presents the development of a pulse-coupled oscillators time synchronization simulator on the OMNeT++ platform for simulating and studying its behaviour and performance in sensor networks. A clock module with configurable phase and frequency noises, and adjustable and higher resolution is developed to mimic various crystal oscillators in embedded systems, for example, the real-time clock. The developed simulator also supports the full functions devices defined by ZigBee protocol, which allows realistic simulation of multi-hop IEEE 802.15.4 wireless networks. Finally, the intensive simulations of classical PCO with the refractory period in IEEE 802.15.4-based WSNs have been carried out to demonstrate the features and benefits of the developed simulator. It is shown that for the non-identical and time-varying PCO clocks in the WSNs, the achieved synchronization will lose gradually, and the time that maintained synchronization depends on the length of refractory period.

Time synchronization for wireless sensor networks (WSNs) has been studied in recent years as a fundamental and significant research issue. Many applications based on these WSNs assume local clocks at each sensor node that need to be synchronized to a common notion of time. Time synchronization in a WSN is critical for accurate time stamping of events and fine-tuned coordination among the sensor nodes to reduce power consumption. This paper proposes a bidirectional, reference based, tree structured time synchronization service for WSNs along with network evaluation phase. This offers a push mechanism for (i) accurate and (ii) low overhead for global time synchronization. Analysis study of proposed approach shows that it is lightweight as the number of required broadcasting messages is constant in one broadcasting domain.

Similar to other cyber infrastructure systems, as wireless sensor networks become larger and more complex, many classic algorithms may no longer work efficiently. This paper presents a wireless sensor network time synchronization model that was initially inspired by synchronous flashing of fireflies. Synchronous flashing of fireflies is an interesting phenomenon that has been studied for decades. A variety of models have been proposed to explain this phenomenon, among which is the pulse-coupled oscillators model that models fireflies as oscillators. The oscillators in such a model interact only through discrete pulses, similar to the flashing of fireflies. In this paper, we propose a new nonidentical linear pulse-coupled oscillators model and use the model to analyze synchronization of pulse-coupled oscillators with different frequencies. The conditions to achieve and maintain synchronization are derived, and then, the results are used to prove that the oscillators in the model can achieve synchronization eventually, except for a set of frequencies with zero Lebesgue measure. Furthermore, through simulations and implementation on a wireless sensor network testbed, we demonstrate that the proposed nonidentical linear pulse-coupled oscillators model can be used in designing lightweight scalable time synchronization protocols for distributed systems.

Impact of time synchronization error on the mode-shape identification and damage detection/localization in WSNs for structural health monitoring

In distributed systems like large wireless sensor networks, dense indoor small cell networks, maintaining each node's logical clock synchronized to a common reference is challenging. Increasing data rates and required cooperation among nodes for high spectral efficiency, further exacerbate the required clock synchronization accuracy. Here, we propose a one way timing information distribution mechanism simplifying existing two way information exchange for clock synchronization. We view the problem at the synchronizing node as a Peer-to-Peer (P2P) continuously running process. A node with stable clock (master) transmits regular packets embedded with transmit time stamps. All synchronizing nodes within the master's communication range can listen and apply receive time stamp. Efficient Passive Minimum Variance Unbiased Estimator (PMVUE) or Passive Maximum Likelihood Estimator (PMLE) are applied to jointly estimate clock offset and skew to correct the local clock. Random component of the link delay is assumed to be Gaussian distributed. Cramer-Rao Lower Bound (CRLB) guarantees that PMVUE and PMLE are efficient. We also propose a less accurate PMLE method, Independent PMLE (IPMLE) to remove dependency on random delay distribution parameter and show through simulations the high accuracy of our algorithms.

Clock synchronization plays a crucial role in Wireless Sensor Networks (WSNs). Assuming that there is no clock skew between sensor nodes, the Maximum Likelihood Estimate (MLE) of clock offset was derived by [1] for clock synchronization protocols assuming exponential random delays and a two-way message exchange mechanism as in TPSN (Timing-sync Protocol for Sensor Networks [2]) or NTP (Network Time Protocol). The MLE is appropriate for the case that the random delays in WSNs are exponentially distributed. However, the performance of the MLE is deteriorated considerably in the case that the underlying distribution is contaminated by or mixed with other distributions. Hence, a robust estimator is needed. In this paper, clock offset estimators based on the robust M-estimation method are proposed and it is shown that the proposed estimators present excellent performance in the mean squared error (MSE) sense under the condition that the underlying distribution is mixed with other distribution.

An efficient clock offset datum switching compensation method for BDS real-time satellite clock offset estimation

For many applications, distributed networks require the local clocks of the constituent nodes to run close to an agreed upon notion of time. Most of the widely used clock synchronization algorithms in such systems employ the sender-receiver protocol based on a two-way timing message exchange paradigm. Maximum likelihood estimator (MLE) of the clock offset based on the timing message exchanges between two clocks was derived in D. R. Jeske, On maximum likelihood estimation of clock offset[IEEE Trans. Commun., vol. 53, pp. 53-54, Jan. 2005], when the fixed delays are symmetric and the variable delays in each direction assume an exponential distribution with an unknown mean. Herein, the best linear unbiased estimate using order statistics (BLUE-OS) of the clock offset between two nodes is derived assuming both symmetric and asymmetric exponential network delays, respectively. The Rao-Blackwell-Lehmann-Scheffe? theorem is then exploited to obtain the minimum variance unbiased estimate (MVUE) for the clock offset which it is shown to coincide with the BLUE-OS. In addition, it is found that the MVUE of the clock offset in the presence of symmetric network delays also coincides with the MLE. Finally, in the presence of asymmetric network delays, although the MLE is biased, it is shown to achieve lesser mean-square error (MSE) than the MVUE in the region around the point where the bidirectional network link delays are symmetric and hence its merit as the most versatile estimator is fairly justified.