Slowly but surely, networking is linking together everything electronic in some way. Whether or not you agree with that, there are benefits to enjoy and money to make. Killer technologies like Ethernet, wireless, and optical are making it happen.
Networking's future appears to stay focused on achieving higher speeds, even though our data rates already do everything we want. The real goal is to get that speed to everyone, and do it in a secure environment.
WANs: After years of build-out in the long-haul telephone and Internet backbone networks, there's more than enough capacity to accommodate needs for the immediate future. The economic downturn has slowed the provisioning of all this capacity. Lots of dark fiber is available to light the future when necessary.
MANs: This is the hot spot in the market today. Metro networks either fail to meet the current speed and capacity needs or don't exist at all, so there are plenty of growth opportunities (see "Optical: Undisputed King Of High-Speed Data Transmission," p. 57).
LANs: Virtually all medium and large enterprises have been fully networked for years. However, there will be an ongoing need for upgrades in service to higher speeds and gradual replacement of older systems. More and more enterprises are adopting wireless networking as well because its declining cost allows them to take advantage of the flexibility it offers.
PANs: We're just now starting to see the effects of personal area networks. Most are wireless, with the Bluetooth star showing the way. It took years for Bluetooth to come into its own, but today it's poised for major growth, so you'll be seeing more of it and such spinoffs as ZigBee. IrDA isn't dead, and the newer inductive wireless networks have great potential in some applications.
Ethernet Everywhere: Clearly, Ethernet is the networking technology making the most impact. Even though it's been around for decades, it has considerable life left (Fig. 1). The original 10-Mbit/s coax systems have mostly been replaced by 100-Mbit/s twisted-pair Fast Ethernet. A newer version handles 1 Gbit/s on CAT5 (Category 5) wiring as a LAN backbone or to the desktop.
The 1- and 10-Gbit/s optical fiber versions of Ethernet will ensure a continuation of this flexible and endurable technology. The 1-Gbit Ethernet (1 GE) configuration is being widely adopted in large LAN backbones and storage-area networks (SANs). Already, 1-GE systems are showing up in new metro networks. No 10-Gigabit Ethernet (10 GE) systems are yet in operation, but these will find applications in large fast LANs and SANs, metro networks, and even WANs replacing Sonet systems. Ethernet has also been blessed as the de facto industrial networking standard too.
What's next for Ethernet, 100 Gbits/s? Not right away, as the technology isn't there yet. Most likely, a 40-Gbit/s version will take advantage of available Sonet OC-768 semiconductor technology and provide some compatibility with Sonet in the metro and long-haul networks.
Additionally, Ethernet has come to dominate wireless networking. Wireless is increasingly becoming the physical layer of choice in many LANs, and the IEEE 802.11b Ethernet has become the de facto standard. The latter has flourished because the Wi-Fi consortium ensures interoperability between many vendors' equipment. Prices have dropped due to competition and more widespread adoption.
Wireless Ethernet also has a life. The recently approved 802.11g standard offers a speed upgrade path for 802.11b to 22 Mbits/s, and even 54 Mbits/s. And, 5-GHz 802.11a standard products are slowly rolling out.
Frontiers For Growth: Aside from opportunities for expansion in the metro network arena, several major sectors will experience growth: mesh networks, home networks, cell-phone networking, broadband, UWB, and security.
The classical network topology is the mesh, where all nodes are connected to each other (Fig. 2). Even though they're very expensive to implement, there's a growing interest in mesh networks because they provide network redundancy and thus QoS. Further, mesh networks scale more easily than other topologies, so many local carriers and metro network vendors are seriously considering them.
Mesh networks also are proposed as the wireless broadband solution in locations where cable and DSL aren't available. Every home wireless node acts as a router, creating a mesh network with each of its neighbors. Such mesh networks provide excellent Internet access and could be a solution for the wireless local loop.
Another trend is the growth in home networks. Wireless dominates with Ethernet 802.11b in the lead, but viable phone-line and power-line network products have finally become available. Affordable access gateways now make it fast and easy to network several PCs at home.
Plus, the new 2.5G cell phones have become nodes in a huge wireless data network, providing e-mail and Internet access. But the 2.5G data services, like GPRS and 1xRTT, aren't selling well. This is not good for forthcoming faster 3G phones. Many believe consumers will prefer a wireless LAN access port to a cell phone for Internet access via their laptop. There's growing consensus that high-speed data access isn't what most cell-phone subscribers need, as the killer application for cell phones remains voice.
Perhaps the networking opportunity of greatest importance and benefit is broadband. Data applications like e-mail and Internet access will continue to dominate the Internet. But as the world moves closer to being a totally networked group, it will bring interactive gaming, more online education, and voice, movies, and music on demand—but only when everyone has an affordable broadband connection.
Currently, less than 10% of U.S. homes have a broadband connection. Statistics say that out of the approximate 10 million current broadband connections, roughly 7% are cable modems and 2.8% are DSL. Less than 2% use some form of wireless/satellite broadband link.
Cable TV's hybrid fiber-cable (HFC) networks are almost everywhere now. This very reliable and affordable technology will keep dominating broadband. Most project cable as the continuing leader in broadband growth.
As for DSL, most of us are impressed, even amazed, that we can achieve DSL speeds of up to 1.5 Mbits/s on the POTS (plain-old telephone system) local loop. But speed is typically much less over longer distances and with more noise. Let's face it, DSL is a dead-end technology. Billions of dollars are being spent to roll out a technology that has no room to grow. We should move on to the better killer technology, optical fiber.
Wireless networking will also benefit from forthcoming ultra-wideband products. In February, the Federal Communications Commission (FCC) approved the use of UWB for networking and in short-range radar applications. UWB wireless LANs should produce data rates of over 100 Mbits/s and at very low power. This will enable inexpensive video links in home networks and even wireless versions of IEEE 1394 and USB 2.0.
Finally, and maybe of greater importance, is the resolution of security issues in Internet and wireless applications. While there are some software and hardware solutions in the form of firewalls, encryption, and more recently biotechnology, we're still waiting for "the" elusive product that will make this happen. The deployment of some applications will continue to be slowed by the technical complexities of the problem, the lackadaisical attitude about security in industry despite terrorist and hacker threats, and the widespread disagreement over which approaches to take. Indeed, security is one of the better opportunities in networking today.