Introduction
Simply stated, the mobile Internet can be thought of as the migration of standard Internet applications and services to a mobile environment. The introduction of mobility, however, raises a number of new considerations: What wireless technology is most appropriate for the provision of Internet services? Is this technology equally appropriate for applications with dissimilar requirements, such as e-mail and video broadcasting? How do we provide ubiquitous Internet services while users move across different locations, where the same wireless service may not be available? Will it be more appropriate to consider several wireless technologies, such as cellular data networks, wireless local area networks (WLANs), wireless personal area networks (WPANs), and so forth? If so, then how do we combine them into a seamless wireless service? How do we optimize the utilization of wireless resources to accommodate as many mobile Internet users as possible? How do we handle the security issues raised by the wireless transmission and possibly by the use of different wireless services provided by different operators?
These are just a few of the questions we must address in our attempt to make the mobile Internet a reality. One important clarification is in order here: The fact that today we can use our laptop or personal digital assistant (PDA) along with a wireless device (e.g., a cellular phone or WLAN adapter) to establish access to our Internet/intranet e-mail or other Internet Protocol (IP) services does not really mean that the mobile Internet is already available. In reality, what is defined as the mobile Internet is far more complex than that. By definition, the Internet is a network of millions of users who communicate by means of standard Internet protocols. Therefore, in the mobile Internet we also need to support millions of always-connected mobile/wireless users. The key phrases we need to bear in mind are "always connected" and "millions of users." To be "always connected" means that we do not have to make a connection before every transaction or after we change wireless service. Our assumption is that we always have connectivity to public Internet services and that we are always reachable through a public IP address, no matter where we are. This calls for extensive mobility management and seamless handover across different wireless networks. And we do need several wireless networks (and several wireless technologies, as we will see later on) because no sole wireless network can provide ubiquitous wireless services, and therefore no sole wireless network can meet the always-connected requirement. A typical mobile Internet user is assumed to move seamlessly between different wireless networks (or even between fixed and wireless networks), which may or may not employ the same radio-access technology. This is schematically illustrated in Figure 1.1. The seamless fashion of movement suggests that our (virtual) connection to the public Internet is transferred from one access network to the other without any actions from the user's side. In effect, this creates a virtual wireless network from the user's perspective that provides ever-present Internet connectivity.
We need also to support millions of users wirelessly connected to the Internet in order to be compliant with the large scale of users supported by the Internet. This creates capacity concerns and better explains why no single wireless network is sufficient. The capacity concerns have direct implications in several design aspects. For instance, the wireless technology needs to be as spectrum efficient as possible; it has to implement a random access scheme that can accommodate a large number of users and degrade gracefully in overload conditions; the IP addressing scheme should be able to support the required user capacity; and the end-toend transport schemes and applications should probably take mobility into account.
It is interesting to note that most research projects nowadays, as well as standardization activities, are moving around such goals. In this context, the correlation between the mobile Internet and the so-called beyond third-generation (3G) technologies is evident. In reality, most of the beyond-3G technologies are tailored to support the requirements for mobile Internet: increased capacity, quality of service (QoS), mobility, security, TCP/IP enhanced performance, and integration of diverse technologies into one ubiquitous virtual network.
No comments:
Post a Comment