NDIS-based virtual polling algorithm for IEEE 802.11b for guaranteeing the real-time requirements

Abstract

This paper focuses on the feasibility of IEEE 802.11b wireless LAN for real-time wireless industrial network. In general, it has been known that IEEE 802.11b is not suitable for industrial networking because its medium access control method is the contention-based CSMA/CA (Carrier Sensing Multiple Access with Collision Avoidance) that exhibits unstable performance under heavy traffic and unbounded delay distribution. As an attempt to circumvent the probabilistic nature of IEEE 802.11b, this paper presents an enhanced four-layer architecture using the Network Driver Interface Specification (NDIS) and a virtual polling algorithm. Also, the enhanced four-layer architecture's performance is compared with that of conventional IEEE 802.11b. Based on the experimental results, it is found that the enhanced four-layer architecture for IEEE 802.11b is a very promising alternative for wireless industrial networking.

Introduction

Currently, mobile devices such as mobile robots, Automated Guided Vehicles (AGVs), and Overhead Hoist Transfers (OHTs) are widely used in many material-handling systems. These mobile devices are essential for the movement of materials or parts, and consequently for the productivity of automated systems. One example of such a system is the 300-mm wafer system in Fig. 1, in which several OHTs are used to transport wafers to various locations in the system [1]. For the system to operate efficiently, many control functions are needed, including functions for task allocation, collision avoidance, and motion control. These control functions require various types of information, such as the locations of individual OHTs from the distributed sensor or bar code system along the rail, job order information calculated by Material Control Systems (MCSs) and job-scheduling systems, and reports from OHTs and process equipment regarding the status of a task. Therefore, there is a need to exchange information among various subsystems, such as MCSs, OHTs, process equipment, and sensor systems. In general, various wire-based protocols such as DeviceNet are used for data exchange among stationary systems such as sensor systems and MCSs. In contrast, mobile systems like OHTs are connected via a wireless communication, such as a Radio-Frequency (RF) modem.

In general, data exchanged on an industrial network can be classified into two groups: real-time and non-real-time data. Non-real-time data do not have stringent time limits on the communication delays experienced during data exchange. By contrast, real-time data have very strict time limits and the value of data is diminished greatly as the communication delay grows longer. This real-time data can be further divided into hard and soft real-time data, depending on the severity of the detrimental effect of delay on the system performance. For example, sensing information that is collected by sensor systems and transmitted to an OHT through an MCS can be regarded as hard real-time data, while the current location of an OHT for task allocation can be regarded as soft real-time data. In addition, the job order message for an OHT is non-real-time data. The other criterion is the periodicity of data generation. For example, data from a sensor for a closed-loop control are periodic data, while a task-completion report is aperiodic in nature. On many industrial networks, these data types share a single network, although they have different communication requirements. That is, the non-real-time data need assurance of delivery without error and duplication, while the real-time data are concerned mostly with the time taken to reach the destination. Therefore, when building an industrial network, the designer must configure the network to satisfy these requirements [2], [3].

To satisfy the real-time requirements, various standard organizations have developed wireless industrial networks since the late 1990s. As a part of the European Strategic Program for Research and development in Information Technology (ESPRIT), the open low-cost time-critical wireless fieldbus architecture (OLCHFA) was initiated, but applicable research results are not yet available [4]. The R-Fieldbus project supported by the European Commission is under development. It selects spread spectrum technology for the physical layer, and the Profibus protocol for the upper layer of the data link layer. The R-fieldbus can satisfy real-time requirements, but a practical application is not yet available [5], [6].

Recently, as an alternative to the wireless industrial network, IEEE 802.11b [7] has attracted some attention because of its simplicity and wide acceptance. Lee et al. [8] employed both wired and wireless networks to integrate mobile devices for material handling. Ye et al. [9] introduced the applicability of IEEE 802.11b in a network-based control system. In addition, Tang et al. [10] reported an application of for IEEE 802.11b for an IC manufacturing system.

However, IEEE 802.11b is unsuitable for industrial networking because the Medium Access Control (MAC) method of IEEE 802.11b is contention-based Carrier Sensing Multiple Access with Collision Avoidance (CSMA/CA), and its performance becomes unstable with heavy traffic and an unbounded delay distribution. Hence, in order to accept IEEE 802.11b as a wireless industrial network, an algorithm that can prevent the collision of frames in the network is necessary [11].

As an alternative for the uncertainty of IEEE 802.11b, this paper presents an enhanced four-layer architecture using the Network Driver Interface Specification (NDIS) [12], [13], [14] and a virtual polling algorithm. The NDIS used here was developed by Microsoft and 3Com for programming the device driver layer, which handles the transmission and reception of packets, by activating the data link layer or hardware device directly. That is, if the NDIS is applied to IEEE 802.11b, the user can handle all the raw packets received from other Network Interface Cards (NICs) and can transmit any packet to an “own” NIC. With these NDIS characteristics, it is possible to eliminate frame collisions and solve the uncertainty of IEEE 802.11b by means of the enhanced four-layer architecture using the NDIS with a virtual polling algorithm.

This paper is organized into five sections including this introduction. Section 2 gives a brief overview of IEEE 802.11b and the NDIS, and Section 3 presents the enhanced four-layer architecture using the NDIS and virtual polling algorithm. An experimental test bed for the suggested enhanced four-layer architecture was implemented and is compared with the performance of the conventional IEEE 802.11b network in Section 4. Finally, a summary and conclusions are presented in Section 5.