KR20180124487A - Node-Scheduling Control Method for Wireless Sensor-Actuator Networks with Time Delay Constraints - Google Patents

Abstract

The present invention relates to a node scheduling control method of a wireless sensor-actuator network (WSAN) having time delay constraint. The node scheduling control method minimizes energy consumption in a dynamic WSAN and, at the same time, enables time delay recognition data propagation inspired by an epidemic theory. The node scheduling control method of a WSAN having time delay constraint includes: a step of receiving a packet of a source node in a specific node; a step of measuring a packet time delay based on transmission time information included in the received packet; a step of updating a propagation delay time based on the measured packet time delay; a step of delivering a packet to the updated propagation delay time; and a step of determining an activation or sleep state of the specific node.

Description

TECHNICAL FIELD [0001] The present invention relates to a node-scheduling control method for a wireless sensor-actuator network having a time delay constraint,

The present invention relates to a node scheduling control method of a wireless sensor-actuator network having a time delay constraint, and more particularly to a dynamic sensor-actuator network (WSAN) The present invention relates to a node scheduling control method for a wireless sensor-actuator network having a time delay constraint to provide a node scheduling control method capable of propagating time delay recognition data inspired by a node.

Wireless sensor networks (WSN) connect devices that can detect and monitor various physical phenomena. Generally, a WSN includes a number of small sensor nodes running on a battery. The WSN system has several applications that focus on goal tracking, environmental monitoring, traffic control, and scientific inquiry. In recent years, actuator nodes have been introduced that perform various tasks, and their sensors and actuators have been able to form networks such as wireless sensor-actuator networks (WSANs) (Non-Patent Documents 1 and 2). In the WSAN, the sensor collects information about the physical environment, while at the same time the actuator makes appropriate decisions and takes appropriate action. WSAN applications include disaster relief, intelligent buildings, military surveillance systems, health monitoring, and infrastructure security. Depending on the control application in the WSAN, it is necessary to react quickly to the sensor input. Specifically, mission critical applications, such as emergency situations and structural applications, require applications-specific functionality and reliable performance assurance. Also, the sensor data must be valid at the time of taking action since it takes action after detecting the environment. Therefore, an efficient and reliable data transmission system must operate in a timely and accurate manner. Unfortunately, the existing medium access control (MAC) protocol is primarily designed for energy savings (see Non-Patent Documents 3 to 6). However, energy savings result in additional end-to-end time delays, which makes it difficult for WSAN systems to meet reliable communication requirements in a dynamic network change environment. Therefore, design for reliable node scheduling with dual goals of energy savings and limited propagation time delay for WSAN systems is important.

Biologically inspired modeling techniques have attracted considerable interest in computer networks as a new alternative to realizing a computer network that is robust, scalable, and adaptable but maintains individual simplicity (see Non-Patent Literatures 7-11) ). Particularly, the present invention relates to an immune system (see Non-Patent Documents 12 to 15) and insect populations (see Non-Patent Documents 16 to 20), active substances and inhibitor systems (Non-Patent Documents 21 to 24) 25-30), have been applied to wireless network system design.

Among biologically inspired models, epidemic-based communication models provide an effective means of transmitting data over wireless networks. Here, the infection propagation process is related to the message transfer between nodes (see Non-Patent Documents 31 to 39). The similarities between dissemination of information in wireless networks and propagation of communicable diseases within communities are clear. When a node receives a piece of information and stores it, it can say "infected." Otherwise, it can be considered "infectable." The epidemic is considered a chain reaction of infectious diseases. (See Non-Patent Document 32). These researchers have focused on efficient propagation of individual node observations at all nodes, and have found that data descriptors that eliminate redundant transmissions in the network (refer to non-patent document 33). [0004] The Firecracker protocol (see Non-Patent Document 34) is a method of transmitting data through a wireless network to data And the broadcast principle is used in combination to propagate rapidly. 35) has been proposed to propagate and manage code updates as quickly as possible to all nodes in the network. One of the key contributions of this protocol is to suppress the rate of broadcast to limit transmission between neighboring nodes, Quot; polite gossip " that utilizes the in-tuning of a node, which provides a mechanism for determining when a node propagates code. [0005] Deluge (see non-patent reference 36) Still other researchers have proposed a gossip-based approach in which each node determines the delivery of a message to another node based on a certain probability (see Non-Patent Document 37). If the source has very few neighbors, the information may be hidden before reaching all nodes. In a modified protocol, for example called GOSSIP1 (p, k), each node sends information with probability 1 during the first k moves until gossip persists probability p, . Middleware has been proposed to provide controlled infectious mode propagation using information propagation techniques that regulate the process according to desired confidence (see Non-Patent Document 38). A multi-rumor overwriting (MRO) model has also been proposed (see Non-Patent Document 39). This model describes the propagation and interaction of periodically sensed data, which is distributed using gossip, without causing significant interference to the network.

(Non-Patent Document 1) O. Akan, I. Akyildiz, and V. C. Gungor, "A Real Time and Reliable Transport Protocol for Wireless Sensor and Actor Networks," IEEE / ACM Trans. on Networking, vol. 16, no. 2, pp. 359-370. 2008. (Non-Patent Document 2) H. Salarian, K. Chin, and F. Naghdy, "Coordination in wireless sensor-actuator networks: A survey," Journal of Parallel Distributed Computing, vol. 72, no. 7, pp. 856-867, 2012. (Non-Patent Document 3) W. Ye, J. Heidemann, and D. Estrin, " An Energy Efficient Macro Protocol for Wireless Sensor Networks, " IEEE INFOCOM, pp. 1567-1576, 2002. (Non-Patent Document 4) D. Moss and P. Levis, " BoX-MACs: exploiting physical and link layer boundaries in low power networking, " Tech. Rep. 08-00, Stanford University, 2008. (Non-Patent Document 5) H. Byun and J. Yu, "Adaptive duty cycle control with queue management in wireless sensor networks," IEEE Trans. Mobile Computing, vol. 12, no. 6, pp. 1214-1224, 2013. (Non-Patent Document 6) A. Bachir, M. Dohler, T. Watteyne, and K. K. Leung, " MAC essentials for wireless sensor networks " IEEE Communications Surveys & Tutorials, vol. 12, no. 2, pp. 222-248, 2010. (Non-Patent Document 7) F. Dressler and O.B. Akan, " A survey on bio-inspired networking, " Computer Networks, vol. 54, no. 6, pp. 881-900, 2010. (Non-Patent Document 8) T. Nakano, " Biologically inspired network systems: A review and future prospects, " IEEE Trans. Systems, Man, and Cybernetics, Part C: Applications and Reviews, vol. 41, no. 5, pp. 630-643, 2011. (Non-Patent Document 9) C. Charalambous and S. Cui, " A biologically inspired networking model for wireless sensor networks, " IEEE Network, vol. 24, no. 3, pp. 6-13, 2010. (Non-Patent Document 10) K. Leibnitz and M. Murata, "Attractor selection and perturbation for robust networks in uctuating environments," IEEE Network, vol. 24, no. 3, pp. 14-18, 2010. (Non-Patent Document 11) C. Cheng, C. K. Tse, and F. C. Lau, "A bio-inspired scheduling scheme for wireless sensor networks," IEEE Vehicular Technology Conference, pp. 223-227, 2008. (Non-Patent Document 12) T. Bokareva, N. Bulusu, and S. Jha, "SASHA: Toward a self-healing hybrid sensor network architecture," Proc. 2nd IEEE Workshop Embed. Netw. Sens., Pp. 71-78, 2005. (Non-Patent Document 13) U. Aickelin, J. Greensmith, and J. Twycross, "Immune system approaches to intrusion detection-A review," Proc. 3rd Int. Conf. Artif. Immune Syst. Lect. Notes Comput. Sci., Vol. 3239, pp. 316-329, 2004. (Non-Patent Document 14) C. Lee and J. Suzuki, "An immunologically-inspired autonomic framework for self-organizing and evolvable network applications," ACM Trans. Auton. Adap. Syst., Vol. 4, no. 4, 2009. (Non-Patent Document 15) B. Atakan and O. B. Akan, "Immune system based distributed node and rate selection in wireless sensor networks," Proc. 1st IEEE / ACM Int. Conf. Bio-inspired Models Netwo., Inform. Comput. Syst., Pp. 1-8, 2006. (Non-Patent Document 16) A. Montresor, H. Meling, and O. Babaoglu, " Messor: Load-balancing through a swarm of autonomous agents, 1st Int. Workshop Agents Peer-to-Peer Comput., Pp. 125-137, 2002. (Non-Patent Document 17) N. Bean and A. Costa, " An analytic modeling approach for network routing algorithms that use ant-like mobile agents, " Int. J. Comput. Telecommun. Netw., Vol. 49, no. 2, pp. 234-268, 2005. (Non-Patent Document 18) G. W. Allen, G. Tewari, A. Patel, M. Welsh, and R. Nagpal, "Firey-inspired sensor network synchronicity with realistic radio eects," Proc. ACM Conf. Embed. Netw. Sens. Syst., Pp. 143-153, 2005. (Non-Patent Document 19) J. Degesys, I. Rose, A. Patel, and R. Nagpal, "DESYNC: Self-Organizing Desynchronization and TDMA on Wireless Sensor Networks," Proc. 6th Int. Conf. Inform. Process. Sens. Netw., Pp. 11-20, 2007. (Non-Patent Document 20) S. Barbarossa and G. Scutari, " Bio-inspired sensor network design, " IEEE Signal Process. Mag., Vol. 24, no. 3, pp. 26-35, 2007. (Non-Patent Document 21) F. Dressler, "Self-organized event detection in sensor networks using bio-inspired pro-motors and inhibitors," ACM / ICST Int. Conf. Bio-Inspired Models of Network, Information and Computing Systems (Bionetics 2008), 2008. (Non-Patent Document 22) F. Dressler, "Bio-inspired feedback loops for self-organized event detection in SANETs," IEEE / IFIP International Workshop on Self-Organizing Systems (IWSOS 2008), vol. LNCS 5343, pp. 256-261, 2008. (Non-patent Document 23) K. Hyodo, N. Wakamiya, E. Nakaguchi, M. Murata, Y. Kubo, and K. Yanagihara, "Experiments and Considerations on Reaction- Int. Symp. World of Wireless, Mobile and Multimedia Networks (IEEE WoWMoM 2007), pp. 1-6, 2007. (Non-Patent Document 24) G. Neglia and G. Reina, "Evaluating activator-inhibitor mechanisms for sensor coordination," IEEE / ACM Int. Conf. Bio-Inspired Models of Network, Information and Computing Systems (IEEE / ACM BIONETICS 2007), 2007. (Non-Patent Document 25) N. A. Monk, J. A. Sherratt, and M. R. Owen, "Spatiotemporal patterning in models of juxtacrine intercellular signaling with feedback," Institute for Mathematics and Its Applications, vol. 121, pp. 165-193, 2001. (Non-Patent Document 26) J. R. Collier, N. A. Monk, P. K. Maini, and J. H. Lewis, "Pattern formation by lateral inhibition with feedback: A mathematical model of delta-notch inter-cellular signaling," Theoretical Biology, vol. 183, no. 4, pp. 429-446, 1996. (Non-Patent Document 27) C. Charalambous and S. Cui, " A biologically inspired networking model for wireless sensor networks, " IEEE Network, vol. 24, no. 3, pp. 6-13, 2010. (Non-Patent Document 28) I. Wokoma, L. Shum, L. Sacks, and I. Marshall, "A biologically-inspired clustering algorithm dependent on spatial data in sensor networks," Proc. 2nd Eur. Workshop Wireless Sens. Netw., Pp. 386-390, 2005. (Non-Patent Document 29) K. Leibnitz and M. Murata, "Attractor selection and perturbation for robust networks in uctuating environments," IEEE Network, vol. 24, no. 3, pp. 14-18, 2010. (Non-Patent Document 30) M. Paone, A. Cucinotta, A. L. Minnolo, L. Paladina, A. Puliato, and A. Zaia, "A bio-inspired distributed routing protocol for wireless sensor networks" IEEE Int. Conf. Distributed Computing Systems, pp. 247-255. 2010. (Non-patent Document 31) Pradip De and Saal K. Das, Y. Gao, and D. Zuo, "Epidemic models, algorithms and protocols in wireless sensor and ad-hoc networks," John Wiley & Sons, doi: 9780470396360. ch3, 2008. (Non-Patent Document 32) J. Kulik, W. Rabiner, H. Balakrishnan, " Adaptive protocols for information dissemination in wireless sensor networks, " Proc. the Fifth ACM / IEEE Int. Conf. Mobile Computing and Networking, pp. 174-184. 1999. (Non-patent Document 33) S. S. Kulkarni and M. Arumugam, " INFUSE: a TDMA based data dissemination protocol for sensor networks, " Technical Report MSU-CSE-04-46, Dept. of Computer Science, Michigan State University, 2004. (Non-Patent Document 34) P. Levis and D. Culler, " The recracker protocol, " Proc. the 11th ACM SIGOPS European workshop, 2004. (Non-Patent Document 35) P. Levis, N. Patel, D. Culler, and S. Shenker, "Trickle: a self-regulating algorithm for code propagation and maintenance in wireless sensor networks," Proc. the 1st conf. Symp. Networked Systems Design and Implementation, vol. 1, pp. 2-2, 2004. (Non-Patent Document 36) J. W. Hui and D. Culler, " The dynamic behavior of a data dissemination protocol for network programming at scale ", Int. Conf. Embedded Networked Sensor Systems, 2004. (Non-Patent Document 37) Z. J. Haas, J. Y. Halpern, and L. Li, "Gossip-based ad hoc routing," IEEE / ACM Trans. Networking vol. 14, no. 3, pp. 479-491, 2006. (Non-Patent Document 38) M. Musolesi and C. Mascolo, " Controlled epidemic-style dissemination middleware for mobile ad hoc networks, " 3rd Ann. Int. Conf. Mobile and Ubiquitous Systems, pp. 1-9, 2006. (Non-Patent Document 39) G. Williamson, D. Cellai, S. Dobson, and P. Nixon, "Modeling Periodic Data Dissemination in Wireless Sensor Networks," UKSim European Symposium on Computer Modeling and Simulation, pp.499-504, 2009 . (Non-Patent Document 40) R. M. Anderson and R. M. May, "Infectious diseases of humans: dynamics and control," Oxford University Press, 1992. (Non-Patent Document 41) M. E. J. Newman, " The structure and function of complex networks, " SIAM Review, vol. 45, pp. 167, 2003.

However, the algorithms inspired by the infectious diseases of the prior art have a common weakness. They do not allow control over information propagation. It is therefore difficult to be certain that a system can reach the desired performance level for a particular application using fixed parameters under dynamic network conditions.

In addition, existing algorithms are mainly based on experimental results rather than analytical models. For this reason, existing biologically inspired systems have not been theoretically analyzed in terms of system stability.

SUMMARY OF THE INVENTION Accordingly, the present invention has been made in view of the above problems, and it is an object of the present invention to provide a method and apparatus for minimizing energy consumption in a dynamic wireless sensor-actuator network (WSAN) The present invention provides a node scheduling control method for a wireless sensor-actuator network having a time delay constraint for providing a node scheduling control method capable of propagating time delay recognition data.

Another object of the present invention is to design an algorithm that is inspired by infectious diseases for node scheduling considering proper spreading in a dynamic WSAN environment and to control the node state according to the time delay requirement of the application field, And to provide a node scheduling control method of a wireless sensor-actuator network having a time delay constraint.

According to another aspect of the present invention, there is provided a method for controlling node scheduling of a wireless sensor-actuator network having a time delay constraint, the method comprising the steps of: (a) receiving a packet of a source node at a specific node; (b) measuring a packet time delay based on transmission time information included in the received packet; (c) updating the propagation delay time based on the measured packet time delay; (d) delivering the packet with the propagation delay time updated in the step (c); And (e) determining an active or sleep state of the particular node.

The transmission time information may be time stamp information.

In the step (c), the electronic delay time is updated to a maximum value between the measured packet delay and the previously stored propagation delay time while the propagation delay time value is stored.

In the step (e), the active or sleeping state of the specific node is determined every time it is turned on every time slot (I _m ) for every time slot.

In the step (e), when activated, it remains activated for a time interval T _alpha and then enters a sleep mode.

In the step (e), if new data is received during the activation, the data of the specific node is subjected to the broadcast processing.

According to the present invention, a theoretical model based on infectious disease system is presented, and the performance of the WSAN system is automatically minimized in a dynamic network condition, and the desired propagation time delay is provided.

In addition, according to the present invention, there is also an advantage in providing insight into the performance of the proposed system using analytical formulas to obtain network responsiveness and steady state.

1 is a conceptual diagram of a wireless sensor-actuator network to which the present invention is applied,
FIG. 2 is a flowchart illustrating a node scheduling control method of a wireless sensor-actuator network having a time delay constraint according to the present invention;
Figs. 3 (a), 3 (b) and 3 (c) are explanatory diagrams of the influence of Na on the propagation rate, the propagation time delay,
FIG. 4 is a diagram illustrating a basic operation of the node scheduling algorithm proposed in the present invention,
5 is an exemplary network topology used in the simulation,
6 shows the basic parameters used in the simulation,
Figure 7 is a graph of the time behavior of Na with time delay requirements,
Figure 8 shows an average Na graph that varies with delay time requirements,
9 (a) and 9 (b) are examples of energy consumption and propagation time delay of the proposed protocol,
FIG. 10 is a graph showing energy consumption and propagation time delay results compared with the protocol for the number of nodes of 100. FIG. 10A shows the protocol proposed in the present invention, FIG. 10B shows the SMAC protocol, FIG. protocol,
FIG. 11 is a graph showing energy consumption and propagation time delay results compared with the protocol for the number of nodes 200. FIG. 11A shows the protocol proposed in the present invention, FIG. 11B shows the SMAC protocol, FIG. protocol.

Hereinafter, a node scheduling control method of a wireless sensor-actuator network having a time delay constraint according to a preferred embodiment of the present invention will be described in detail with reference to the accompanying drawings.

1 is a schematic diagram of a wireless sensor-actuator network to which a node scheduling control method of a wireless sensor-actuator network having a time delay constraint according to a preferred embodiment of the present invention is applied.

In FIG. 1, reference numerals 11 to 10 + N denote sensor nodes for collecting information on the surrounding environment and transmitting the collected environment information to the actuator 31 through the wireless network 20.

Reference numeral 30 denotes a target device to be automatically controlled by the user and the actuator 31 controls the control target device 30 based on the environment information acquired from the sensor nodes 11 to 10 + And scheduling the sensor nodes of the wireless sensor-actuator network having a time delay constraint.

FIG. 2 is a flowchart illustrating a node scheduling control method of a wireless sensor-actuator network having a time delay constraint according to the present invention. FIG. 2 illustrates a process of performing node scheduling in the actuator 31, (B) measuring a packet time delay based on the transmission time information included in the received packet (S20); (c) (S30) of updating the propagation delay time based on the measured packet time delay, (d) transferring the packet to the propagation delay time updated in the (c) step (S40), and (e) And determining an active or sleep state (S50).

The node scheduling control method of the wireless sensor-actuator network having the time delay constraint according to the present invention configured as described above will be described in detail as follows.

Before describing the present invention in detail, the infectious disease theory applied to the present invention will be described as follows.

The epidemic theory explores the epidemiology of the spreading of infectious diseases in the population (see non-patent reference 40). The epidemic model models the process of transmission of the infection. Two of the simplest models for explaining this propagation process are Susceptible Infected Recovered (SIR) and Susceptible Infected Susceptible (SIS) models (see Non-Patent Document 41). In SIR, a vulnerable person is infected, and after a period of time called the infection period, the person is recovered and is immunized for further infection. In SIS, the person becomes vulnerable again after recovery.

In the general SIR model, the epidemiology of infection (I (t), vulnerability (S (t)) and recovered person (Z (t)) in total population (N (t) Can be explained by the same differential equation (see Non-Patent Document 40).

Where β and γ denote infection rate and infection removal rate, respectively. The SIR model explains that the proportion of these vulnerable people is reduced and the rate of infection is calculated using infectious β , the number of vulnerable subjects S (t), and the number of infected persons I (t). The total number of nodes is maintained for the time interval considered. The SIS model does not have a fully recovered person (Z (t)), and the recovered people are again included as vulnerable after the initial recovery.

In epidemiology, the network topology is used to study the spread of disease infections among the population, nodes represent people, and corners represent social contact. The WSAN system is identical to the SIS model of the epidemic, which can be in only two states (vulnerable or infected), and the state change is the result of infection among individuals. Network protocols inspired by existing epidemics have focused on modifying SIS models to achieve more realistic behavior of wireless networks (see Non-Patent Documents 31-39).

The present invention proposes a time delay recognition data propagation system inspired by the infectious disease theory described above. For example, we propose a node scheduling method that translates disease propagation methods into data propagation models and then determines the state of each node considering the time delay constraints and energy efficiency for a specific application.

Node scheduling control is the most important in the WSAN, which has a high level of impact across all other aspects of sensor network operation. Without any specific node scheduling, the node can not deliver data to the actuator quickly, or consumes a significant amount of energy while keeping the radio on for long time intervals. To demonstrate the effect of node scheduling on time delay, the inventor introduces _a parameter Na, which is the average number of nodes to be turned on during a particular activation period. If N _a is too low, the packet is hidden before it is propagated through the entire network.

Figure 3 (a) shows the influence of N on _a transmissibility (delivery ratio). This result indicates that the minimum value of N _a is needed to ensure a certain level of confidence. In the ensuing simulation, the inventor has set the minimum N _a to 30, which achieves a transmission rate of 90% or more. It should be noted that the minimum N _a can be automatically controlled based on the transmittance.

Figure 3 (b) shows the average propagation time delay for various Na values. _A small time delay when N _a is low is due to low transmittance. If N _a is large enough to achieve _a high transmission rate, the time delay decreases with increasing N _a .

Figure 3 (c) also shows that there is a linear increase in energy consumption when N _a increases. This result shows that the time delay is controlled by the number of nodes activated in the network. With this motivation, the inventor designed a time delay aware node scheduling controller with time delay constraints specific to the application field.

The present inventors have considered a WSAN in which an actuator node is deployed in a multi-hop topology and one or more source nodes (sensor nodes) propagate messages to all actuators. By rewriting a set of nodes with N = {1, 2, ..., N}, each node can exist in only one of two states: active or sleep. The node turns off the radio when it is sleeping and turns on the radio to exchange information packets when activated.

4 is a diagram illustrating a basic operation of the model proposed in the present invention.

Initially, all nodes were periodically turned on with period I _M conditions. If, while the node is on, when it receives new data, the node transmits its data to the neighbor with a calculated probability, which is after it has been turned on the period I _m (I _m <I _M ). Whenever a node is turned on at every time slot I _m for every time slot, it determines where to go to the next active or sleep mode based on the proposed epidemic formula. When activated, the node remains active for a time interval T _a , and then goes into sleep mode. During activation, if the node receives new data, the node broadcasts the data.

The present invention enhances the reliability of the WSAN so as to ensure performance required for application in a dynamic environment. Node scheduling proposed a discrete time formula in which the control algorithm is updated every control period τ. Therefore, the time is divided into [n, n + 1], n = 0, 1, ... at the same time interval τ.

The present inventor adopted the SIS model of the known infectious disease theory and made a discrete dynamic model for the proposed method as shown in the following Equations (2) to (4).

Where N _s (n) and N _a (n) are the number of sleeping and active nodes in timeslot n, α is the infection rate, β is the control parameter, and N _t is the total number of nodes. Equation (2) - Equation (4) means that the number of active nodes is calculated by the dynamic infection degree?. Existing infectious disease algorithms used predefined infectious doses and were not used to adjust for infection according to environmental changes. In order to solve this problem, the present inventors propose a node scheduling method for controlling a value according to network conditions as well as a desired performance for an application field.

Before designing the node scheduling method, the present inventor again expresses the propagation time delay and the application delay time requirements as d (n) and dr, respectively.

As shown in FIG. 2, when a source node (sensor node) transmits a packet, it includes a time stamp indicating a transmission time. Each node stores the propagation delay time value in the table. The initial propagation delay time value is 0, and is reset to the initial value every control period. When the node (actuator) receives the packet in step S10, the node measures the packet time delay from the source node to itself in step S20. Next, in step S30, the stored propagation delay time is updated to a maximum value between the previously stored propagation delay time and the measured packet time delay from the received packet. In step S40, the node forwards the packet to the newly stored propagation time delay. Using the propagation delay time, the node scheduling method is modeled by the following equation (5).

Where ε (<< 1) is a positive constant and η is a control parameter that must be selected. The function Φ is the increment function in the interval [-d _M , ∞], where d _M is the upper bound of the delay time requirement. The combined model of Equations (2) - (5) increases the value of alpha when the measured time delay exceeds the time delay requirement, leading to the newly activated node of all the nodes. Instead, if the measured time delay is less than the time delay requirement, the value of alpha decreases and causes an increase in the number of nodes in the sleep state.

Using the Na values determined by Equations (3) and (5) above, each node determines, during the next type slot, whether the node will be active or sleeping. To embody specific node scheduling, the inventor introduces a uniformly distributed random value ω in the [0, 1] interval. Each node independently generates a random number. If the ratio of the active node to the total node (Na / Nt) is smaller than?, The node goes into a sleep state. On the other hand, if the activated ratio is larger than?, The node becomes active during the next type slot. If the propagation time delay exceeds the required time delay, the values of α and Na are increased, further increasing the probability of activation. This causes reduction of the delay time and more frequent packet transmission, but if the propagation delay time is smaller than the required time delay, the values of α and Na decrease and the sleep time increase leads to energy saving. If the node does not receive any packets during the k consecutive timeslots while it is active, the measured time delay value d is used as 0 and Na is decreased according to [Equation 3] - [Equation 5]. If d is still zero and Na reaches the minimum Na, the node returns to its initial state and is turned on in IM every time.

As a result, the present inventors provide a way for the proposed method to efficiently control the infection rate a and the number of activated nodes Na for a given application requirement level and network conditions, The node scheduling is adaptable according to the node schedule.

In order to simplify the above equations, the present inventor limited the function Φ (σ (n)) to the first order as shown in the following equation (6).

가정하에 N_as와 αs를 상기 [수학식 3]과 [수학식 5]에 따른 정상상태 값이라고 하고, 포화 한계를 제거하면 상기 [수학식 3]은 하기 [수학식 7]과 같이 다시 쓰여 진다.That Nt is approximately constant during total infection

Assuming that N _as and a s are the steady state values according to the above equations (3) and (5) under the assumption that the saturation limit is removed, the above Equation (3) is rewritten as Equation (7) .

In Equation (7), a normal point as shown in the following Equation (8) is obtained.

It can be observed that the number of activated nodes changes according to the delay time requirements according to Equation (5) - (8). As the time delay requirement becomes larger (larger value), the value of αs becomes smaller and the value of N _as becomes smaller. On the other hand, as the delay time requirement becomes stricter (smaller value), the value of? S becomes larger and the value of N _as becomes larger.

이라고 가정한다. 여기서 d_s는 측정된 시간지연의 정상 상태이다.

를 만족하는

를 고려하면 상기 [수학식 8]은 다음과 같이 제한된다.

. Where d _s is the steady-state of the measured time delay.

Satisfy

(8) is limited as follows.

Since 0 <N _as <N, we derive β condition as follows.

Further, the condition of eta in (8) is derived as follows.

를 고려한다.As follows

.

Then the time delay in steady state is derived as follows.

In the steady state, the time delay range is determined by the delay time requirements dr and σ ^* . If δ ^* (> 0), the range of Nas and ds and the control parameters β and η is automatically determined according to [Equation 9] - [Equation 13].

In order to verify the stability of the proposed system, if x = Na / N, then Equation (2) - Equation (5) can be simply written as:

Displays tr as the time constant (rise time) of the network to reach steady state. In Equation (14), the time constant tr is given as follows.

Here, αs is determined as follows.

According to Equation (15) - (16), the responsiveness of the proposed system greatly varies according to the delay time requirement dr. That is, as the delay time requirement becomes stricter (smaller), the value of alpha becomes larger and causes a short rise time to reach a steady state. On the other hand, the longer the time delay requirement is, the longer the rise time for the network becomes. This means that it takes a longer time to reach the Nas new state of the activated node part. Also, the value of [alpha] increases with [sigma] ^* , which results in a shorter convergence time. That is, the larger the value of [sigma] ^* , the wider the desired range of the delay time, and thus the electronic time delay can converge within a desired range in a shorter time. On the other hand, if the range of [sigma] ^* is small, the time required to converge to the desired time delay range of [Equation 13] becomes long.

라 하면, 다음과 같은 선형 시스템을 얻을 수 있다.To verify the stability of the system, the nonlinear system was approximated as described in equation (14) through linearization for the mean steady state point.

, The following linear system can be obtained.

이다. 제어기가 안정화될 수 있도록 [수학식 17] 의 특성 다항식은 단위 원내에 모든 제로(zero)를 가져야 한다. 따라서 제안된 시스템은 제어 파라미터가 다음 관계를 만족한다면 점근적으로 안정(asymptotically stable)하다.And

to be. To be able to stabilize the controller, the characteristic polynomial of (17) must have all zeros in the unit circle. Therefore, the proposed system is asymptotically stable if the control parameters satisfy the following relation.

If the parameter is set within the range of [Equation 18] - [Equation 19], the proposed system converges in the steady state of [Equation 13]. Therefore, the trajectory of the nonlinear state equation will behave like a stable node.

The present inventors have simulated the proposed system.

The simulation was performed using a simulator written in C ++. The simulator captures real-life events such as carrier detection, backoff, and collision. At the start of the simulation, the nodes were randomly placed within the simulation area, and the source nodes were centrally located. An example topology can be seen in FIG. Using the basic parameters shown in FIG. 6, the maximum hop distance at the source node is mostly 4 to 6. Once deployed, the node switches between the active mode and the sleep mode to start the operating cycle. At one point, the source node begins to periodically generate packets. Packets are marked with a sequential number so that the node does not receive or forward duplicate packets. The goal of the source node is to send packets to all other nodes (actuators) in the network within a given time delay requirement.

Two different medium access control protocols, SMAC (see Non-Patent Document 3) and BoX-MAC (see Non-Patent Document 4), are executed in the simulator. SMAC is a MAC with a synchronized duty cycle in which all nodes are simultaneously turned on and off. Whereas the BoX-MAC is a MAC with an unsynchronized duty cycle where the node is turned on and off at another point in time. In general, broadcast-based applications are well suited to SMAC, and BoX-MAC is well suited to unicast-based applications.

The proposed protocol is implemented on SMAC. Time synchronization is performed by the SMAC and the proposed protocol determines if the node is actually awake at the beginning of the duty cycle period. Two different schemes were implemented for comparison, and their performance was evaluated. The first was the SMAC protocol with a fixed duty cycle period. Although SMAC does not automatically control based on time delay requirements, manual regulation of the duty cycle can achieve a balance between energy consumption and packet delay. The source broadcasts the packet for data propagation, and the other node broadcasts the packet once again when it receives the packet. The second is a gossip protocol that runs on BoX-MAC (see Non-Patent Document 39). BoX-MAC is a unicast-based approach to gossip and is used as a MAC protocol because of its better performance when run on BoX-MAC. When a node generates or receives a packet, the node simultaneously forwards the packet to c among its neighbors. C is called the gossip tx count and the period between the c transfers is called the gossip tx interval. If C is too small, the packet is not propagated to the entire network. In the simulation, c is set to the minimum value with a transmission rate exceeding 90%. Similar to SMAC, the duty cycle of a node is varied to study the effect on energy consumption and packet time delay.

Three evaluation methods were used for performance evaluation.

Average Transmission Rate: The number of nodes that received packets divided by the total number of nodes averaged over all packets sent by the source node.

Average propagation time delay: the time between the source node that sent the packet and all the nodes that receive the packet on the network, averaged over the entire packet (ignoring the node that did not receive the packet)

Average energy consumption: The amount of charge consumed by a node within a second, averaged over the entire simulation time. The unit is [mA-s].

The basic parameters used in the simulation are listed in FIG. These parameters are used unless otherwise specified. Except for the graph showing the behavior of the proposed protocol over time, one point in the graph is the average for 500 simulation runs with various network topologies and various random seeds.

The experimental results are as follows.

In Fig. 3 (a), the influence of Na on the transmission rate, propagation time delay, and energy consumption has been described. As discussed in the previous section, this result indicates that although Na reduction reduces energy consumption by balancing propagation time delay, the minimum Na must be set to maintain a transmission rate above a certain threshold. The minimum Na was chosen as 30, while the maximum Na was equal to the total number of nodes.

In FIG. 3 (b), it can be observed that there is a range in which the balance between propagation time delay and energy consumption can be controlled. For example, for 100 nodes, the propagation delay time can be controlled between 15 and 30 seconds. Even if the delay time requirement is longer than 30 seconds, Na can not be further reduced because of the transfer rate. Even if the delay time requirement is less than 15 seconds, Na can not be increased significantly beyond the total number of nodes. If the time delay requirement is out of range, the duty cycle length (wakeup period) can also be controlled, so the delay time requirement falls within a controllable range.

7 shows the time behavior of Na according to the proposed scheme for the delay time requirements of 15 seconds, 20 seconds and 25 seconds, respectively. The number of nodes in this experiment is 100. It can be observed that Na converges to various values based on the time delay requirement. That is, Na converges higher as the required time delay is shorter. Since the smaller Na consumes less energy, it can be said to control the propagation delay time to save energy if it meets the application requirements. In addition, the results prove the time constant of (15) that the α value increases as the time delay requirement becomes stricter and conversely the rise time to reach steady state becomes shorter.

Next, the average Na of the proposed scheme according to the delay time requirements varying from 0 to 60 seconds is shown in FIG. It can be appreciated that the average number of activated sensor nodes is controlled by the time delay requirement. The longer the time delay required, the greater the average Na. Similar to previous experiments, the results show that there is a range that can control the balance between propagation time delay and energy consumption. Using 100 nodes, the change in Na occurs when the time delay requirement is between 10 and 30 seconds. With 200 nodes, this range is between 10 and 50 seconds.

9 (a) and 9 (b) show the energy consumption and the maximum delay time of the proposed scheme. Within the controllable range, we can see that the proposed method can control the parameters to save energy while meeting the latency requirements. The dotted line in the graph is the case where the propagation time delay is equal to the delay time requirement. It can be observed that a delay time requirement can be achieved within a controllable range.

When combining the results of FIG. 8 and FIG. 9, it is shown that the more the margin of delay time requirements are, the lower the Na value, which reduces energy consumption. On the other hand, as the latency requirement becomes more stringent, Na increases, which results in more frequent packet transmissions to meet latency requirements.

If the time delay requirement is set outside the controllable range, the propagation time delay and average energy consumption are fixed at one of the end points. Compared to the results of Figs. 9 (a) and 9 (b), higher node densities can achieve longer network lifetimes because of lower energy consumption per node, if they can withstand longer delay times.

10 is energy consumption and propagation time delay compared according to the protocol on the basis of the number of nodes 100. 10 (a) shows energy consumption and propagation delay time according to the protocol proposed in the present invention, FIG. 10 (b) shows energy consumption and propagation delay time according to the SMAC protocol, and FIG. 10 Gossip, energy consumption and propagation delay time according to Box-MAC protocol.

11 is energy consumption and propagation time delay compared according to the protocol, based on 200 nodes. 11 (a) shows energy consumption and propagation delay time according to the protocol proposed in the present invention, FIG. 11 (b) shows energy consumption and propagation delay time according to the SMAC protocol, and FIG. 11 Gossip, energy consumption and propagation delay time according to Box-MAC protocol.

Simulation results show that the proposed scheme has superior performance over existing schemes and achieves high transfer rate, guaranteed delay time and energy saving.

While the present invention has been particularly shown and described with reference to exemplary embodiments thereof, it is to be understood that the invention is not limited to the disclosed exemplary embodiments, but, on the contrary, It is obvious to those who have.

11 ~ 10+: Sensor node
20: Wireless network
30: Control target device
31: Actuator

Claims (8)

1. A method for controlling node scheduling in a wireless sensor-actuator network having a time delay constraint,
(a) receiving a packet of a source node at a particular node;
(b) measuring a packet time delay based on transmission time information included in the received packet;
(c) updating the propagation delay time based on the measured packet time delay;
(d) delivering the packet with the propagation delay time updated in the step (c); And
(e) performing a step of determining an active or sleep state of the particular node,
Wherein the scheduling of the nodes in the wireless sensor-actuator network has a time delay constraint.
The method of claim 1, wherein the transmission time information of the step (b) uses time stamp information.
The method of claim 1, wherein the step (c) updates the electronic delay time to a maximum value between the measured packet delay and the previously stored propagation delay time while storing the propagation delay time value. A method for controlling node scheduling of a wireless sensor - actuator network having constraints.
The method of claim 1, wherein step (e) comprises determining the active or sleep state of a particular node each time it is turned on in a time interval (I _m ) for every time slot. Node scheduling control method.
The method of claim 1, wherein step (e) comprises activating a time interval (T _alpha ) and entering a sleep mode if the activation is active. .
The method according to claim 1, wherein, in the step (e), when new data is received while the node is activated, the node performs a multicast process on the data.
The method of claim 1, wherein the step (e) introduces a uniformly distributed random value ω in a [0, 1] interval, each node independently generates a random number, Characterized in that if the ratio (Na / Nt) is less than [omega], the node enters a sleep state and the node remains active during the next type slot if the activated ratio is greater than [omega] A method for controlling node scheduling in a sensor - actuator network.
The method as claimed in claim 1, wherein the step (e) controls the infection rate a and the number of activated nodes Na by using the following formula to determine the active or sleep state of the node: A method for controlling node scheduling of an actuator network.

Where N _s (n) and N _a (n) are the number of sleeping and active nodes in timeslot n, α is the infection rate, β is the control parameter, and N _t is the total number of nodes.

Images