problem A: If you use a non directional antenna (easiest to set up, no alignment issues) you are then presented with the amount of processing needed to weed out signal from reflections - it is enormous. Your antenna also has no gain - a big problem with lossy low power MMDS or LMDS systems. No signal = lots of noise = low bandwidth or high error rate.
I assume you're referring to using decision feedback equalization (DFE) to mitigate the Inter-symbol interference (ISI). This technique is absolutely unreasonable in non-LOS environments. The processing power required is exponential with bandwidth.
It isn't, however, too difficult to frequency multiplex (FDM) a large number of narrowband signals into the same wide-band. As bandwidth increases, more narrow-bands are added. As long as these bands are small enough, ISI is decreases to the point that no extra compensation is required. Voila: linear complexity increase with bandwidth.
Now, you're probably thinking "what about guard bands? They'll eat your bandwidth for lunch." Nope. Using orthogonal FDM (OFDM), the signal is coded such that guard bands aren't necessary. If you really want the nitty gritty, check out this page
If vendors insert such wiretapping capabilities into their routers' source, such a "feature" will almost certainly be optional.
As voice over IP and other routed protocols becomes more prevalent, it is possible that the government will require the ability to wiretap these communications. Industries under such fiat will have to chose the wiretap-enabled version.
Large IP networks providing secure VPN services for businesses will almost certainly chose the software without such wiretapping capacity.
Vendors will produce whatever their customers will buy. If any customer requires a backdoor (via government fiat or otherwise), every vendor will have that feature faster than you can say 'eavesdrop'.
Lucent is the only company I know of capable of routing OC-48 IP at line speed. They have imminent plans to route OC-192 IP at line speed, and I expect them to release it soon. Some can route gigabit ethernet, but routing OC-48 and up is where the men are separated from the boys.
Cisco's performance OC-48 card (not the old time-to-market card) for the GSR's does line rate ip quite beautifully.
Lucent has one other competitor in the terabit routing space, and that is Juniper.
Juniper's box is *not* a terabit router. It has 40 gbps of bandwidth. There is no such thing as a terabit router yet. Nexabit's OC192 is not real OC192. It's basically Nexabit's OC48 engine (which sucks) combined with the Lucent mux/optics package. Nothing special, and not too surprising. Just lets you run the 4*OC48 straight into the router rather than using an external optics package. The wait for true OC192c continues.
Lucent just recently finished their purchase of Nexabit, a private datacomm company. Nexabit currently is shipping the NX64000, a true 6.4 terabit IP router mentioned at the bottom of the article.
I don't know where you got the 6.4Tb number comes from. Let's see...we have 16 slots per box. Put an OC192 in each slot. That's 10 gigs per slot. Hrmm... that's 160 gigs. Even if you could put *10* OC192s in a slot (which you can't), that's only 1.6Tb/s. Yeah, it's a no-brainer all right. Just give me a Cisco 12012. At least they're honest and call their box a gigabit router.
Slashdot Mirror
It isn't, however, too difficult to frequency multiplex (FDM) a large number of narrowband signals into the same wide-band. As bandwidth increases, more narrow-bands are added. As long as these bands are small enough, ISI is decreases to the point that no extra compensation is required. Voila: linear complexity increase with bandwidth.
Now, you're probably thinking "what about guard bands? They'll eat your bandwidth for lunch." Nope. Using orthogonal FDM (OFDM), the signal is coded such that guard bands aren't necessary. If you really want the nitty gritty, check out this page
If vendors insert such wiretapping capabilities into their routers' source, such a "feature" will almost certainly be optional.
As voice over IP and other routed protocols becomes more prevalent, it is possible that the government will require the ability to wiretap these communications. Industries under such fiat will have to chose the wiretap-enabled version.
Large IP networks providing secure VPN services for businesses will almost certainly chose the software without such wiretapping capacity.
Vendors will produce whatever their customers will buy. If any customer requires a backdoor (via government fiat or otherwise), every vendor will have that feature faster than you can say 'eavesdrop'.