Why is Setting up WiFi so Inefficient

A couple of months ago, I moved to a new apartment. I had AT&T in my old place and transferred it to my new one. I couldn’t make a reservation for the new place until the previous tenant canceled service because he had AT&T. The installer had to do a lot of searching for connections. Why wouldn’t they have that in the systems since the previous person had AT&T. The whole process was very inefficient. Dime supposedly lives in Japan so it’s hard to blame it on him.

2+

Users who have liked this post:

  • avatar
  • avatar

Author: Richard Easton

Co-author of GPS Declassified: From Smart Bombs to Smartphones.

14 thoughts on “Why is Setting up WiFi so Inefficient”

  1. I worked in telecom for a while.  My observation is that the telecom system is a jumbled mishmash of decades of systems patched together and made interoperable with baling wire, string, and  good intentions.  Miraculously it hangs together, but the IT infrastructure really shows the patched-together nature of the system.

    I imagine this is an easy problem to fix in one system, if 6 other systems could incorporate some patches, causing another 50 systems to need fixes, etc. So, they just add one more system around the whole mess and hope it works for a while longer.

    I worked overseas in a developing country for a couple of years, and the system we created was super-clean, because they literally cleaned out all the wires and equipment from their 50’s-era existing system and replaced it all with a single-vendor fiber optic system.  Life was so easy, simple, and efficient!  But I expect that given another 50-60 years it will be a jumbled up mess.

    4+

    Users who have liked this comment:

    • avatar
    • avatar
    • avatar
    • avatar
  2. AT&T lies.

    They advertise a bandwidth, but the advertised bandwidth is only available to the customers near one of their hubs.   The farther away from the hub you are, and the more customers connected to your hub, the farther below the advertised service you fall.

    Last year I was getting ready to move from a popular suburb out to the edge of the ruburbs.   The distance is only eight miles but there is a big difference in development density.   Two customer service techs and their sales person all said that it would be no problem for me to keep the level of internet service I had been paying for.   I thought that was unlikely, so I repeated the question with each of them.

    Of course, after we moved in to the new house, it turned out that our neighborhood has its own hub, but that the hub is connected to AT&T via copper, not fiber, so 5 Mbps is the max available.   I grumped, but rather than get entangled with Comcast, I kept U-verse.

    It turns out that I do not notice the difference in the lower level of service.   Which tells me how much lower than advertised was the service I had been getting before I moved.

    2+

    Users who have liked this comment:

    • avatar
    • avatar
  3. MJBubba:
    Of course, after we moved in to the new house, it turned out that our neighborhood has its own hub, but that the hub is connected to AT&T via copper, not fiber, so 5 Mbps is the max available.

    Fiber vs. copper has little to do with it. All connections are multiplexed at some point so the bandwidth is shared. We used to have fiber, now have copper. I just tested the down BW and got 70 Mbps. However, I’m sure when everyone in the neighborhood is streaming video this evening it will be more like 10 Mbps. The point is, copper can deliver plenty of BW.

    2+

    Users who have liked this comment:

    • avatar
    • avatar
  4. Richard Easton:
    The installer had to do a lot of searching for connections. Why wouldn’t they have that in the systems since the previous person had AT&T. The whole process was very inefficient.

    I had always thought that there must be some kind of master database (probably distributed) which kept track of all of the hard wired connections in the telecom network.  After all, every one of those terminal blocks have numbers on each of the terminals and the wires are colour-coded.  It was a great enlightenment when I discovered in the 1970s that there is nothing of the sort.  When they hook up a new line, they attach it to a free terminal pair on the block and then hook up a “warbler” to the line (I don’t know the proper name—I just made that up).  Then they go from junction box to junction box, following the warble, until they get to where they need to connect it to the central office.  Only then is the end-to-end connection recorded.  When I went to have a leased line installed here in Switzerland for Internet connectivity in 1993, it was the same thing: the warbler and the test set was used to trace the connection all the way to the central office.

    2+

    Users who have liked this comment:

    • avatar
    • avatar
  5. drlorentz:

    MJBubba:
    Of course, after we moved in to the new house, it turned out that our neighborhood has its own hub, but that the hub is connected to AT&T via copper, not fiber, so 5 Mbps is the max available.

    Fiber vs. copper has little to do with it. All connections are multiplexed at some point so the bandwidth is shared. We used to have fiber, now have copper. I just tested the down BW and got 70 Mbps. However, I’m sure when everyone in the neighborhood is streaming video this evening it will be more like 10 Mbps. The point is, copper can deliver plenty of BW.

    That may be so, but even AT&T is not willing to advertise any higher level of service in my neighborhood than 5 Mbps.   They told me the limiting factor is the copper.   I don’t necessarily believe anything they say; but that was the excuse they gave.

    0

  6. drlorentz:

    MJBubba:
    Of course, after we moved in to the new house, it turned out that our neighborhood has its own hub, but that the hub is connected to AT&T via copper, not fiber, so 5 Mbps is the max available.

    Fiber vs. copper has little to do with it. All connections are multiplexed at some point so the bandwidth is shared. We used to have fiber, now have copper. I just tested the down BW and got 70 Mbps. However, I’m sure when everyone in the neighborhood is streaming video this evening it will be more like 10 Mbps. The point is, copper can deliver plenty of BW.

    It all depends on how you count bandwidth.  Usually it refers to a maximum rate from the ISP to your equipment.  Some technologies (such as cable modem) share the transport, so more connections tend to push down bandwidth.  Other technologies (such as DSL) don’t share the transport from the ISP to the customer.  Cleverly, DSL uses the same wire as your telephone, but transmits its data at a high frequency unused by voice communications.  That’s why there’s a filter that is put on the phones at the house.

    Of course, upstream congestion applies to all equipment, so any particular site can be swamped when many people are hitting it.

    Not wanting to boast (<– actually not true!), but check out these numbers.  Ironically this nets out to be $10/month cheaper than I was paying for 125M copper connection.

    2+

    Users who have liked this comment:

    • avatar
    • avatar
  7. MJBubba:

    drlorentz:

    That may be so, but even AT&T is not willing to advertise any higher level of service in my neighborhood than 5 Mbps.   They told me the limiting factor is the copper.   I don’t necessarily believe anything they say; but that was the excuse they gave.

    I would believe them.  Most likely you have DSL service, which directly depends on the length of the wire from your house to the switching station.  It sounds like you’re on the far end of that.  I had a connection like that at my old place, so you have plenty of sympathy from me.

    They have plans to eventually switch everything to fiber (which has a lot lower maintenance cost, since they don’t get corrosion problems where wires connect), but that could take eons to roll out everywhere. Ironically there are lots of places in the Silicon Valley with terrible connection rates, as it’s so expensive to build out infrastructure there.

    0

  8. John Walker:

    Richard Easton:
    The installer had to do a lot of searching for connections. Why wouldn’t they have that in the systems since the previous person had AT&T. The whole process was very inefficient.

    I had always thought that there must be some kind of master database (probably distributed) which kept track of all of the hard wired connections in the telecom network.  After all, every one of those terminal blocks have numbers on each of the terminals and the wires are colour-coded.  It was a great enlightenment when I discovered in the 1970s that there is nothing of the sort.  When they hook up a new line, they attach it to a free terminal pair on the block and then hook up a “warbler” to the line (I don’t know the proper name—I just made that up).  Then they go from junction box to junction box, following the warble, until they get to where they need to connect it to the central office.  Only then is the end-to-end connection recorded.  When I went to have a leased line installed here in Switzerland for Internet connectivity in 1993, it was the same thing: the warbler and the test set was used to trace the connection all the way to the central office.

    I had the chance to visit an old class 5 switching station… that’s the one where every home’s wires travel to.  It was before digital cameras, I’m sorry I couldn’t take some pictures.

    There was a big concrete pipe (“sewer pipe”, it seemed to me) that had a 3-foot bundle of bound cables running into the building.  Those were the wires that went to about 4,000 (I think?) homes.  Each pair of wires is connected to the switch.  When I worked at NEC in the early 90s, our biggest class 5 switch was in Hong Kong and had 2 million NEC V-30 CPUs (one for each subscriber), and 16 NEC S-20 CPUs to do the switching.

    Each phone has a dedicated relay and CPU, so it was incredibly relaxing to be in the room… it sounded like thousands of people knitting as the relays turned off and on as people picked up phones.

    The battery room was amazing too.  Hundreds of 5-gallon (or so?) open-topped glass jars containing battery acid, with the anode/cathode rods sticking in them.  You could see the bubbles of hydrogen being released as they operated.  Everything was attached by uninsulated solid copper bus bars a couple of inches square that ran at 480V.  The technicians emphasized that it would be a lot of hassle for them if we were to reach up and kill ourselves by touching them.

    2+

    Users who have liked this comment:

    • avatar
    • avatar
  9. I think it is the World Wide Wifi Conspiracy that is keeping Richard Easton from good service. They know how important he is and whoever controls Richard Easton controls the world. Also because it is just plain fun causing him trouble. (Try it! You’ll see.)

    1+

    Users who have liked this comment:

    • avatar
  10. 10 Cents:
    I think it is the World Wide Wifi Conspiracy that is keeping Richard Easton from good service. They know how important he is and whoever controls Richard Easton controls the world. Also because it is just plain fun causing him trouble. (Try it! You’ll see.)

    I have heard of this conspiracy, and judging by the many members who identify their web sites with its initials it seems quite extensive.

    Sorry Richard Easton, it seems there will be no WiFi for you!

    2+

    Users who have liked this comment:

    • avatar
    • avatar

Leave a Reply