Abstract: Combining mesh and wireless sensor networks can compensate for the insufficient data processing capabilities of wireless sensor networks, improve the utilization rate of sensor data, and increase data value. However, the combination of mesh and wireless sensor networks faces a series of problems such as connectivity, scalability, and task scheduling. This paper proposes a combination framework to realize the integration of wireless sensor networks and mesh, and analyzes the problems encountered in the process of constructing the combination framework and the solutions. Keywords: Wireless Sensor Networks; Grid; Integration Framework [b][align=center]The Key Issues on Integration of Wireless Sensor Networks and Grid Feng Xiufang, Liu Baodong[/align][/b] Abstract: The integration of wireless sensor networks (WSN) and grid can complete the lack of WSN's data processing capacities and improve the sensor data utilization ratio and value. However, there are several issues such as connection, scalability, and task scheduling problems when grid communicating with WSN directly. An integration framework is presented to integrate the WSN and grid. The key issues when constructing the integration framework are presented, and the resolving methods are also mentioned. Keywords: Wireless Sensor Networks (WSN); Grid; Integration Framework 1. Overview Advances in microelectronics, embedded systems, and other technologies have driven the rapid development of wireless sensor network technology. Wireless sensor networks are now applied in many fields such as environmental and biological monitoring, industrial monitoring, and military security monitoring. By deploying a large number of sensor nodes in the monitoring area, the physical world can be precisely measured, improving the quantity and quality of real-world data required by applications and reducing monitoring costs. Wireless sensor networks have become a new computing platform that can seamlessly connect the digital world and the physical world; it consists of a series of sensor nodes, each of which has environmental perception, data processing and wireless communication capabilities. Sensor nodes are characterized by battery power, limited computing and storage capabilities and low communication bandwidth, which limits their ability to process and utilize the data they obtain [1]. Now, grid technology, which has the characteristics of high-speed computing capabilities, massive storage capabilities and high-speed communication bandwidth, has become a standard way to solve large-scale distributed and heterogeneous resource sharing in dynamic virtual communities. Combining wireless sensor networks and grids can effectively make up for the shortcomings of wireless sensor networks and has the following advantages: (1) A large amount of data sensed by wireless sensor networks can be processed by grids. The computing and storage resources of grids can process, analyze and store the large amount of data collected by wireless sensor networks. (2) Data obtained by a wireless sensor network can be used by multiple grid applications at the same time. Data obtained by the same wireless sensor network can be used by multiple applications at the same time through the grid platform, making the use of sensor data more convenient and improving the data utilization rate. (3) New knowledge of wireless sensor network data can be obtained by using grids. In the grid, data mining, data fusion, distributed database and other technologies can be used to process the data and obtain new knowledge of sensor data. 2. Related work HourGlass[2] is a combination scheme of grid and wireless sensor network. HourGlass is mainly composed of three parts: data collection network (DCN), sensor access point (SEP), application access point (AEP). DCN is composed of an Internet interconnected system, which can discover, filter and query multiple wireless sensor networks. SEP can map the data requirements of the application to the operation on the underlying wireless sensor network, or route the data on the wireless sensor network to the data collection network (DCN). AEP is the connection system of the application to DCN, which maps the application's request to the DCN-based service for processing. SensorGrid[3][4] is a composite system that combines wireless sensor network and grid. SensorGrid adopts a distributed network structure, which consists of sensor nodes, middle layer and decision-making layer. The system mainly considers distributed data fusion, distributed processing, network collaboration and other issues, and can perform data fusion, transaction monitoring and classification, distributed decision-making and other work. 3. Key Issues in Combining Wireless Sensor Networks and Mesh Networks Wireless sensor networks and mesh networks are two very different networks, differing in physical layer, communication protocols, application protocols, and other aspects. The network connectivity, scalability, and task scheduling issues encountered during the combination of wireless sensor networks and mesh networks need to be addressed using the combination framework proposed in this paper. (1) Different Network Connectivity Issues: In wireless sensor networks, the interconnection between sensor nodes is achieved through low-bandwidth, high-latency, and unreliable wireless networks. Wireless connections between sensor nodes are subject to interruptions due to environmental noise and signal attenuation. In mesh networks, the interconnection of various devices is achieved through fast and reliable wired networks. The combination framework needs to address the unpredictable network interruptions and communication delays in sensor node wireless communication. (2) Protocol Mapping between Wireless Sensor Networks and Mesh Networks: Mesh communication uses standard Internet protocols, such as TCP/IP and HTTP. Wireless sensor network communication typically uses proprietary protocols, especially MAC protocols and wireless routing protocols, which are mostly proprietary. Due to the limited computing and storage capabilities of sensor nodes, they cannot use Internet protocols. Therefore, the combination framework needs to effectively map the network communication protocols used in the mesh to the nodes of the wireless sensor network. In addition, the OGSA standard of the mesh is based on Web Services, which uses technologies such as XML, SOAP and WSDL. It is impractical for sensor nodes to package sensor data into XML format and publish it as a mesh service. It is necessary to combine the framework to map sensor data to mesh services. (3) Scalability The combined framework needs to dynamically add wireless sensor networks to the mesh without changing the overall structure. It should be able to connect multiple wireless sensor networks at the same time and can be easily integrated with the mesh's computing and storage resources so that users can use multiple wireless sensor networks transparently. (4) Energy management Sensor nodes are powered by batteries and their power is usually not replenishable. Energy management is a very important issue in wireless sensor networks. From the perspective of the combined framework, the availability of sensor nodes depends not only on their current load status, but also on their remaining energy. The combined framework should be able to provide adaptive energy management services so that applications using wireless sensor networks can find a balance between sensor node operation and power usage. (5) Task scheduling The task scheduling of sensor nodes in wireless sensor networks should take into account energy consumption and available sensor resources. Meanwhile, wireless sensor networks are data-centric networks, and effectively utilizing the sensor data collected by sensors is also a very important task when scheduling tasks. When multiple wireless sensor networks exist in the combined framework, the scheduling process is required to make full use of various types of data. (6) System Security The data sensed by wireless sensor networks is often very important and requires confidentiality. No data theft or malicious modification is allowed. Mesh resources also require access by certified individuals and service providers. In the mesh, authentication and authorization mechanisms are used to ensure the legitimate identity of visitors and achieve secure access to mesh resources. Wireless sensor networks use node authentication, sensor data encryption, and secure MAC protocols to ensure power saving and effective security of sensor data. In order to ensure the security of both the mesh and wireless sensor networks, the combined framework needs to organically combine mesh security technology and wireless sensor network security technology to ensure the security of the entire system. (7) Robustness Sensor nodes use battery power and communicate with unreliable wireless communication networks, which may cause the sensor tasks running on the sensor nodes to fail. In order to prevent the sensor tasks on the sensor nodes from failing, the combined framework should support task replication and migration. In this way, if some sensor nodes fail, the sensing task can be quickly migrated from the failed sensor node to the normal node. If there are enough sensing resources, the sensing task can also be replicated. In this way, the failure of some nodes will not affect the execution of the entire sensing task. Finally, if the sensing task is interrupted, it should be able to restart from the point of interruption after the system recovers. (8) Quality of Service (QoS) QoS determines whether the system can provide effective sensing resources and services. QoS parameters can specify the sensor nodes, storage space, communication bandwidth, power consumption, and other indicators used by the mesh sensing task. By using these indicators, the robustness of the sensing task can be increased, and the impact of node failure and communication interruption can be avoided. To meet the different QoS requirements of the framework, the QoS requirements specified at the higher level will be mapped to the QoS parameters at the lower level. When the sensing task requires multiple different sensor resources, the sensing resources need to be reserved in order to achieve the required QoS. 4. Wireless Sensor Network and Mesh Integration Framework The wireless sensor network and mesh integration framework can enable multiple wireless sensor networks to access the mesh and provide a unified mesh service. The framework mainly consists of three layers: wireless sensor network access layer, task management layer, and service management layer. The entire system framework is shown in Figure 1. [align=center] Figure 1 Wireless Sensor Network and Mesh Integration Framework[/align] (1) Wireless Sensor Network Access Layer: The main function of this layer is to enable seamless access of multiple wireless sensor networks, abstract the wireless sensor networks, and make the upper layers see a consistent data layer. This layer mainly completes tasks such as network protocol conversion, mesh API mapping, access of multiple wireless sensor networks, security assurance, and task robustness. (2) Task Management Layer: The main function of this layer is to rationally schedule multiple data fusion tasks. This layer mainly completes tasks such as rational allocation of data processing tasks and rational scheduling of multiple sensing tasks. (3) Service Management Layer: The main function of this layer is to manage the wireless sensor network and form and manage wireless sensor network services. This layer mainly completes tasks such as wireless sensor network energy management and quality of service control. 5. Conclusion The integration of wireless sensor networks and mesh can effectively compensate for the problem of insufficient power-saving data processing capabilities of sensors. This paper proposes an integration framework to realize the integration of wireless sensor networks and mesh, and discusses the key issues that need to be solved in the integration process. By utilizing a combined framework, problems such as network interconnection and task scheduling can be effectively solved, enabling seamless integration of wireless sensor networks and meshes, and increasing the processing capabilities of sensor data. However, current combined frameworks have some shortcomings, such as not considering the special handling of mobile sensor nodes and not being able to improve the scheduling methods for sensor nodes, which are areas that need to be improved in the future. References [1] Li Lian, Zhu Aihong. Research on positioning technology in wireless sensor networks [J]. Microcomputer Information, 2005, 21 (9-1): 133-135 [2] Mark Gaynor, Steve Moulton. Integrating Wireless Sensor Networks with the Grid [J]. IEEE INTERNET COMPUTING, 2004, 7: 32-39. [3] Chen-Khong Tham, Rajkumar Buyya. SensorGrid: Integrating Sensor Networks and Grid Computing [J]. CSI Communications, 2005, 7 (29): 24-29. [4] Hock Beng Lim, Yong Meng Teo, Protik Mukherjee etal. Sensor Grid: Integration of Wireless Sensor Networks and the Grid [J], Local Computer Networks, 2005, 91-99.