Random Wireless Blogging from Matthew Glenn Category Archives

March 21, 2007

Making Sense of My Insensitivity and Sensor (Topologies)

Yesterday I got into the elevator on the ground floor of building 14 on the Cisco San Jose campus. I parked my car and made my way to the elevator bank. The elevator arrived, I hit the #4 (for the floor that I work on), and the doors began to close. Right before they were completely shut, someone stuck their hand in, the doors opened (I was bummed out cause I had waited a while), and a person came in the elevator car with me.

Then he hit the number 2.

Now - obviously I could have taken the stairs up to the fourth floor, but it is a long way up... but one freaking flight of stairs??? I couldn't help myself...

"Dude," I asked the guy looking really concerned, "Are your legs okay?"

"Yeah, they're fine," he answered.

"So why didn't you take the stairs?"

He was incredulous...

"That is the most insensitive thing I have ever had someone ask me," he said as the doors opened and he left the elevator.

This exchange got me to thinking about sensors, and what will be the best way to deploy them...

There has been a lot of discussion regarding Wireless Sensor Networks (commonly called WSNs). The promise of sensors everywhere will spawn applications that will make the world an even more efficient place - I'm not going to write about some of these applications.

Given my 'insensitivity', I thought I would write about what architecture WSNs might improve their overall sensitivity I.e. how well they communicate the information the sensors are recording.

Like all networking technologies, we ultimately design around constraints. In the case of WSNs here are a few constraints:

A) Most likely they will operate in unlicensed spectrum (unless a service provider decides to get into the managed WSN business).

B) Many of the devices being sensed will be remote, and will not have access to power, therefore power is a concern.

C) The sensors have to be cheap - especially if they are to proliferate. I once conducted a study and learned that people were willing to spend 1/15th the price of a device on an active RFID tag to track it. So if you want to track a 500 dollar item, then you better have a tag that costs <35 bucks.

Now here are a couple of variables:
1) Required throughput - sometimes this might be high, sometimes it will be a few bytes of information once a day.

2) Technology - Could be WIFI... could be Zigbee... could be the player to be named later...

3) Business criticality - A reporting sensor on a heart rate monitor vs. a sensor on a remote laser printer definitely has different levels of criticality.

So given the above... here is what I think...

I know of two basic architectures for WSNs: Star and Mesh.

A Star topology has each remote sensor connecting to a single hub - much like most Ethernet devices are connected to a switch.

A mesh topology allows each WSN node to 'self organize' into a mesh. These nodes find a central hub (which may be multiple hops away. Therefore information a pressure sensor may take multiple hops through intermediate nodes before finding a central hub.

Ultimately a WSN is deployed get information out of sensors. While the reliability requirements may vary from application to application, no one wants to invest in a wireless technology that performs spuriously. Therefore people should design for success and reliability rather than the cool factor.

In my mind, knowing how complex mesh is, deploying a star topology is probably the right choice - especially because you will likely have time sensitive applications dispersed with less critical applications. I know that all of the sensor companies will talk about how well their mesh works, but is this technology for technology’s sake? Are you willing to bet your WSN something that is neat and possible or something that is reliable?
I'd love to hear what you think.

Did this make me more sensitive?

Posted by Matt Glenn at 11:53 AM Permalink | Comments (2) | TrackBacks (0)

March 14, 2007

Tax Evasion and Matt (TEAM)

It is tax time again. Time to go visit your accountant, buy some tax software, or just do it online.

It is during this time of year that I usually take up new hobbies - anything to keep me away from the ritual of tax preparation. My family has a $10 opt-in pool on when I will announce to them what I'll be doing that year to avoid getting my taxes ready. The person who is closest to the date without going over (Price is Right rules) wins the pool.

Here are some examples of past hobbies:

* A trip to Costa Rica to relearn surfing
* Learning to ride a unicycle
* Going on the road with Phish for a month
* My vision quest to Tibet

This year I've taken up a new hobby, visiting Second Life. What's great about wireless is that I can constantly live in second life so that I don't have to live my real one.

In case you aren't familiar with Second Life, it is a virtual online commuity, complete with Avatars (who represents other people in Second Life), an economy, real estate, schools... literally everything that we have in our day to day life. It is a great way to kill time ( and certainly much safer than knife juggling, which was also under consideration for this year's diversion).

Now here is the kicker... with Wireless, I can constantly be in Second Life. I can be on 802.11 at home, flip over to my CDMA connection on the train, and then go back to WIFI at work.

I love it in Second Life... I don't have any work deadlines - and I certainly don't have to fill out any tax forms (I'm sure it is coming some day). I don't have any bosses in second life.

In fact I'm really only filling out this blog so that people at work will think that I'm actually working in the 'real world', when in fact I wrote this Internet Blog from within my Second Life Connection. Does that mean that the real world is now my Third Life, and I really don't have to pay taxes.

Wireless is cool, but there are some ways to improve my experience. Let me know if you want to know what those ways are.

Posted by Matt Glenn at 03:54 PM Permalink | Comments (0) | TrackBacks (0)

March 09, 2007

The Dreaded Car Computer

I’m on a car thing… Specifically how to merge cars and wireless… Wireless enables mobility, and the most mobile thing I have is my car. So bringing the two together, so that I can stay mobile, seems like a pretty good thing.

Willie Nelson once sang, “You are always on my mind…”

Right now I’m singing, “My car is on my mind…”

Here’s why… I have two cars. One looks like this and the other looks like this… The more expensive one is currently in the shop, which means I’m a racking up bills and thinking about my youth.

When I was younger I used to love to give my cars tune-ups, change the oil, and swap the brakes out… that sort of thing. Any of you real ‘gear heads’ out there will agree that there is no feeling (or smell) quite as sweet as dumping a few quarts of syrupy black goodness into the local sewer so it can make its way out to sea.

I used like to sing the song, “Get back to where you once belonged” while I dumped the stuff. I didn’t *want* to dump it. I might have even felt a bit bad, but I wanted to be part of the cool crowd - and they all did it.

Then these new cars came along and everything was computerized and suddenly the sheer joy of working on car was taken away. It was replaced with the ‘secretive’ car computer. You know the one. For example…

You bring the car in to the shop and Joe the mechanic says, “I’ll plug the computer into it and get back to you with an estimate.”

Next thing you know Joe calls up asking you to fill out a credit application because that little pinging is the beginning of a catastrophic nuclear melt down. He’s gonna have to replace complete fuel injection system, spark plugs, rework the pistons, yada yada yada.

You ask Joe, “How do you know? You haven’t even started working on it?”

To which Joe replies, “The computer told me.”

And you feel helpless.

Two weeks later you go to pick up the car and you notice that Joe, who used to drive a rusty blue Nova with a brown Pinto door, is now driving a candy-apple red Ferrari and dressing like Don Johnson on ‘Miami Vice’.

So anyway yesterday I went to my car shop. I got to talking to my mechanic about what is up with my car. He whipped out his car computer and plugged into my car (Note: I didn’t get a chance to read what the computer said).

However I *did* notice that the computer had a connector on it and that connector was plugged into my car.

I asked him what was up with the computer.

He told me that it is special for the type of car that I have. My car computer communicates what needs replacement and he in turn tells me. I asked him how accurate it is and he told me, “Pretty accurate Matt… Pretty accurate.”

I’m not a big fan of proprietary systems. So I’ve come up with a good idea…

Car companies are talking about putting 802.11 into their cars so that when they come back into the dealer the car can tell the mechanic what is going wrong without the car computer. In addition, for home use your car can be seen like an iPOD so that you can download the latest Interpol album, or perhaps pump a little Judas Priest or Led Zepplin into your car system.

Fine and dandy, but here is what I want:

1) I want my wirelessly-enabled car to upload d its information to an access point at home.

2) I want the car computer, which takes in sensor data from the car to upload its information to a web site

3) I want the web site to interpret the information and tell me what is needed to be done. Note: I’m willing to pay a subscription fee for this information.

4) I want to be able to take this info and then see if I can do it myself, or get an estimate from a few mechanics so I can see if I’m getting a good deal.

Next time you have to ‘plug’ something in to exchange data think about it… that interface can and should be wireless. If you have any ideas about that I’d love to hear about it.

Finally, for those of you that made it this far down my little blog and wetting your quill getting ready to write a complaint letter about me dumping crude oil into the gutter. I have a good defense; I only did it after eating three Hershey’s bars.

It was the sugar and peer pressure. I believe that will hold up in a court of law.

Okay… Okay… The truth is that I’ve never dumped motor oil, but it did make you read this blog!!!

Posted by Matt Glenn at 10:37 AM Permalink | Comments (0) | TrackBacks (0)

March 07, 2007

The best wireless protocol for my commute (or how to find a mate at a snail’s pace or 70 Miles Per Hour)

Multiple times per I drive from my home in San Francisco to building 14 at Cisco’s campus in San Jose…It is not a short drive. Like commuters everywhere I listen to the Radio (NPR, Adam Carolla, and Sports Radio - thank you).

But like every other MTV generation Attention Deficit Disorder (ADD). person on the road my mind wanders. Usually I drift off when commercials are on regular radio and NPR is doing a story about endangered earth worms living on a fault in downtown Los Angeles.

It is at times like this that I turn to one of my favorite distractions. Looking at other drivers on the road and trying to figure out something about them. More specifically (and I don’t know how politically incorrect this is) I try to figure which women are single.

So, being a geek-a-zoid, I’ve decided to write about what the best wireless protocol would be for my new web 2.0 wirelessly-enabled social networking idea: Car-Matchster.

Car-Matchster brings singles together during their commute - bringing back the days of our youth... the days when you would innocently cruise up and dow ‘Main’ street at 15 miles per hour looking for that person you had a crush on in your World History class?

Well Car-Matchster brings that magic back, but at speeds in excess of 75 Miles per hour (except during rush hour).

Here’s how it works. A user inputs information about themselves and the type of person that they are looking for into a Car-Matchster console. As someone drives down the road, Car-Matchster propagates that information and information from as many as eight cars away. The Car-Matchster console sifts through incoming personal data and when two people ‘match’ a notification is sent. The notification might read:

“Match Found - Red Toyota Pickup Truck - Brown Haired 37 year-old surfing male”

The two people then begin driving around frantically ‘looking’ for one another. If they find each other and are interested, then each presses an accept button and an email address is exchanged.

So the question is - what is the best protocol to use between the Car-Matchster consoles? For the sake of this discussion let’s discuss 802.11, WiMAX, Bluetooth, Zigbee, cellular, and Passive RFID.

WIFI - WIFI has a solution for automatically forming groups. It’s called Ad-hoc networks. Basically a group of WIFI clients come together, an ad-hoc SSID is advertised, and they begin exchanging information. I like the pervasiveness of WIFI. Many people have WIFI and just about everyone *knows* about it, which is good. My only issue with WIFI AD-HOC networking is that I don’t know of any inherent hop-count metric. While I *want* to know if there is a single person near me, I don’t want to know about someone that is eighteen miles down the road.

WiMAX - If there was a lot of congestion, or there was a need for class of service in my service, then I think that WiMAX would be a good technology. Perhaps I should ad it into the roadmap so that when we put in the ‘car chat’ feature the QoS is there. I’m not sure about the price of the components so I’m not going to move forward with this protocol.

Bluetooth - Listen - I have problems with my Bluetooth headset working every day. I don’t want to bet the future of my company (or my chances of finding a future wife) on a protocol that I can’t get to work reliably. I know that there are Bluetooth fans out there, and the cost of Bluetooth chipsets are cheap, but my little mind hasn’t been pulled through the collective keyhole.

Zigbee - The Zigbee protocol is a lot like Bluetooth, but it also is a bit more “Meshy” i.e. it is designed to work well in mesh applications. That’s why one of its first application is in sensor networks. Zigbee can operate in any one of of multiple bands. It’s data rate is relatively low, but version 1.0 of the product isn’t going to be transmitting too much information. If I can natively get the hop count out of the protocol, then life will be be pretty good….

Cellular - I recently received a letter from my cell provider it began as follows… “Dear Matt, you are our very favorite customer!!! Because of the number of minutes use every month, we would like you to wear this pin, which is the rough equivalent of the American Express Black Card.” Okay I’m kidding, but I pay a lot of money every month and I’m not too sure that this service would work for them considering how mobile and chatty it would be.

Passive RFID - I like passive RFID. I really like it every time I’m in line at the local grocery store and people are putting their groceries on the belt and then fumbling when it is time for them to pay. Time is money…. Time is money… passive RFID is gonna save me time and therefore money. But how will it work in my application? Here is one theory. The car would be equipped with an RFID identifier as well as a reader. When I pass someone our information would be exchanged. Unfortunately I would only get the information of the person I am driving by and not the information from the person two cars over. So I guess that passive RFID won’t work. But that doesn’t mean that I don’t like passive RFID. I love it.

Time to make a decision…

I think that version 1.0 is going to be Zigbee, with a roadmap to WiMAX when voice capabilities are incorporated into the solution.

As far as I can tell, everyone is going to be signing up for Car-Matchster as soon as I’m finished with product development, which in theory could be soon but for two issues:

1) The Department of Transportation sent me a cease and desist note when I first proposed it. It seems that a little issue called ‘negligence’ was mentioned because it would cause accidents.

2) My girlfriend probably wouldn’t appreciate it

Posted by Matt Glenn at 09:19 AM Permalink | Comments (3) | TrackBacks (0)

February 28, 2007

How do wireless and dry cleaning converge?

Despite the fact that my co-workers joke about my wrinkled clothes - I *do* go to the cleaners. In fact, this morning I went to the one up the street from where I live to drop off some of my wrinkled clothes.

"You work at Cisco Systems Matt?" Ahmed, my dry cleaner, asked me. I have been using him for a long time and we call each other by first name.

"Yeah. How'd you know?"

"I found a business card in one of your shirts. What do you know about RFID?"

I paused for a second... I was truly taken aback. Then it dawned on me that you know that something is going to become pervasive when your dry cleaner asks about it.

I didn't want to give up what I know about RFID too easily. Everyone always asks a 'lawyer' friend for legal advise. This was my big moment to be a domain expert and give up free advise, and I didn't want to just give up what I know to easily.

"What else do you know about me from my shirts?" I asked him.

"You are a pretty normal guy," he said, "but you should see some of my other customers. Check this out."

He came from behind the counter and hit the switch that made all of the finished dry cleaning circulate throughout his little shop and then he made it stop.

"Feel this shirt," he said.

I did and it was a very coarse white shirt with a button down collar.

"... and this one," he added.

In the same bundle of finished dry cleaning there was another coarse shirt, but this one was blue.

"Look at these buttons."

I looked and the buttons on the arms were mismatched.

"The guy that owns these shirts drives a Porsche Carrera and I've been cleaning these same shirts for the last two years. He clearly has his priorities in the wrong place.... Now look at this shirt..."

He hit the switch again, stopped it, and pulled out another bundle of shirts. He asked me to feel all of the shirts, which I did and once again they were coarse.

"Now look at the end of the sleeves."

I looked and the buttons were perfectly matched, but there were stains at the very end of the sleeves.

"This man is a doctor Matt and he has to get new shirts constantly. Its the finish of things that you need to look at... "

The world through the eyes of a dry cleaner was truly a different place.

"What do you want to know about RFID?" I asked.

"When can I get it? I want to be able to find my customers shirts and give them better service. I'd like to tell you that my service is perfect, but I send my shirts out to get them done and periodically the company I send them to loses a shirt, and shortly after I lose a customer. So I want to put RFID on people's shirts. Then I want to track them when they go out."

I told him about passive RFID and how cheap the tags are going to become, and how he can pass the cost of the tags on to his customers. I then asked for a discount because I'm a subject matter expert.

He then told me he was going to charge me double because I'm one of the guys making it happen so I should be able to afford it. He then added that the extra money he is going to charge me is going to help him overcome the guilt he feels for not having it right now.

"One more question???" I asked.

"Go ahead."

"Any advise for me?"

"You always ask for medium starch, that's too much. You should have light starch - its more comfortable and looks better."

"Thanks dude."

Posted by Matt Glenn at 11:37 AM Permalink | Comments (0) | TrackBacks (0)

April 23, 2006

Bring on the BER

What the heck is a BER?

A long time ago I was in a group interview for a candidate ASIC engineer. Generally these are *really* interesting interviews since there is often logic questions asked, design questions, plus normal interview questions. The candidate was an absolute powerhouse - he nailed every question without breaking a sweat.

Out of the blue a guy that I worked with asked a question about how do you compensate for photons affecting bits as they go through the ASIC.

For the first time in the interview the guy paused and said, “I don’t know, but I do know that the likelihood of this happening to any packet would be something like 10-400th power.”

For those of you reading from the cheap seats… That is a very low bit error rate.

The Bit error rate for a standard Ethernet cable is considerably higher, something like 10-23rd or something like that.

We who dwell in the wireless world would salivate for such performane… and I mean salivate. That’s cause the air, by its very nature, is a medium that’s prone environmental variables that ultimately impact the throughput and performance of the wireless network.

Obi-one Kenobi, er, Bob Friday, has said that the Bit Error Rate for wireless is about 10-6th. That means that you are going to have a lot more packets with errors in it, and this is one of the reasons why your performance in wireless doesn’t equal your wireline network.

Beyond the inherent unreliable air, you are also competing with other devices that operate in the same spectrum. That doesn’t help either.

There are new technologies that will help compensate, a prime example is MIMO, which uses multiple antennas - just because a signal was weak, or blocked, from one antenna doesn’t mean that the other antenna had the same problem.

What does a high BER mean for you?

Well if you are surfing to ESPN.com to check on the player personnal activities of the Raiders - who they are picking up to compensate for the bad year that they had last year. Chances are that the BER isn’t going to be a problem… That’s the beauty of TCPIP, and the 802.11 MAC. If you are at the edge of a cell, you may see your client connect and disconnect, but that is more because of signal strength.

Hopefully the Raiders won’t have a high BER when making their decisions.

However, if you are make a voice call, and you have an abnormally high BER, then it may sound like your ear is next to a bowl of Rice Krispies… You know.. Snap, Krackle and Pop.

There are no retransmissions for UDP based applications.

We can improve technology, and believe me we afre working on it, but we will never be able to overcome physics.

Posted by at 10:45 PM Permalink | Comments (0) | TrackBacks (1)

April 06, 2006

Wireless - the convergence protocol

I’m sitting on a plane and like any good product manager I’m thinking about my product line - the size, scope and future of it, and I had a thought...

Without a doubt the world as converged on IP as the protocol of choice. In 1991 we had multi-protocol routers, but IP was already eeking out increasing market share because of the good old Internet. Finally web browsers began being bundled with PCs and voila! Everyone had to get IP connected.

Flash forward fifteen years later and we now run voice over IP, data over IP and video over IP (The IP triple play).

Plus there are a myriad of applications that were non-existent before the world coalesced around this protocol. It has transformed the way we communicate in business and at home.

So now, at 36,000 feet, I’m wondering if wireless has the same potential as IP. A very quotable person once said, “wireless is the natural state of communications.”

I think he was right.

The pieces are in place for the world to converge on WIFI. In much the same way that the web browser made everyone have to be connected to the Internet, Centrino has made it so that every laptop that ships is wirelessly enabled.

Let’s go down the triple play checklist:

Data over WIFI
Well chances are when I post this blog I’ll be sending it over a wireless link in my hotel tonight, or perhaps at a T-mobile hotspot, or perhaps I’ll steal some RF from somewhere - either way. I can do it.

Data over WIFI: CHECK!

Voice over WIFI
I routinely use my IP Communicator over my WiFi link from home - it saves Cisco all kinds of money. I can secure the connection using a VPN client. I’ve done it all over the world and it works really well. For those of you that don’t use IP Communicator, I recommend that you use a freeware program like Skype, Yahoo, or Google talk. They all work.

As you’ll quickly see, the challenge is not the single call on an access point, it is multiple calls on an access point. I often use this example in an EBC….

Imagine that everyone in the room whips out their phone to call their significant other at the exact same time, the way that 802.11 works, all of the client devices would run to the bright light (i.e. the AP with the strongest signal in a room). 802.11 by its nature is a socialist protocol, so we would all be granted access to the medium and the voice quality for all of our calls might suffer.

Its for this reason that we introduced Call Admission Control technology, which when coupled with CCX makes for highly reliable, high quality, scalable voice.

Voice over WIFI: CHECK!

Video over WIFI
Can I watch streaming clips of UCLA losing to Florida in the NCAAs? Yes (I’m a UCLA fan if you are wondering). Can I flip channels and have the same reliable, fast channel changing that I get on my TV? No.

This is clearly an area where there is work to be done. Here are some high level thoughts on what it will take:

• More bandwidth
• Better client capacity management
• Continued decrease in cell sizes (coverage area per AP)
• 802.11e clients

Video over WIFI: Uh.. no! But we’ll get there.

There are a multitude of applications that are growing out of the wireless medium, Wifi presence applications, asset tracking, security applications, monitoring applications, and many more. So from an application enablement perspective I think that wireless actually has an advantage over wireline protocols.

But after meandering through this blog, here is what I’ve concluded.

At Layer 3 IP is the convergence protocol of choice.
At Layer 2 WIFI will be the convergence protocol of choice.
At Layer 1 Wireless is already a dominant medium (think cellular), but it will eventually be the same in the enterprise.

Posted by at 09:47 AM Permalink | Comments (1) | TrackBacks (1)