I could have sworn I said wired networks exist.

Did I say they were still the norm?

No I said acording to the fluff I read and the way things read in SR4 wireless networks are the norm now.

Encryption and firewalls get better as hackers do. But thats what Tiered systems and things like vanishing sans are for. Which are noted to all still be in existance and used.

And realy, unless the wired systems are not connected to the matrix at all, they're just as vulnerable as wireless ones.

Good questions:

1. Spread spectrum technology is specificly designed to be resistant to jamming, both incidential and intentional. By spreading your signal out over a wide bandwith, the jammer must overwealm your signal across a substantial fraction of that bandwith. This is especialy true if FEC (foreward error correction) coding is used, since it is possible to reconstruct a transmition with less then the entire message is correctly recieved. (often called "error correction") Some more advanced SS systems can actualy change what spectrum the operate in in response to envromental conditions. I would expect this type of tech to be standard well before 2070. A SS signal can still be jamed, but it requires the jammer to raise the noise floor across significantly across a wide swath of the spectrum. This requires a lot of power (big batteries and big heat sinks.)

2. Sure. SS can be detected by looking for a sudden rise in the EM noise floor across a given spectrum. Somebody fireing up a signal 5 comlink next to the signal scanner might be fairly clear, depending on the "spread" of the signal. The trade off with this type of system is between "Type I" and "Type II" errors. Type I errors are false Negitives. That is, there WAS a signal and the system failed to detect it. Type II errors are false positives, the system sent a warning, but there was no real threat. It is not possible to eliminate either without driving the other to a limit. You could successfully detect 100% of the signals, but your system would be constantly in an alarm state, and therefore useless. With this tradeoff in mind, youi build your system to suit the enviroment you are in.

3. Depends entrely on how the system is constructed. With "frequency hopping" in WWII the allies used 5-20 frequences, shifing a few times a second. "spread spectrum" is not som much shifting and "smearing" it occupies a bandwidth 10-10,000 times as wide as the data requires, but at 1/10-1/10,000th the signal energy at any given portion of the spectrum. Frequency hopping is not used as often as SS, since SS is cheap and more effective in most circumstances. (there are exceptions to this, so nobody get picky)

4. That's a tough one to answer. If they were all operateing in the same spectrum (and there is no reason they woudl have to be) then they would raise the aparent noise floor seen by the scanner by a log sum function. (my Information theory book is at home, so I am working from memory here) Clearly they would be MORE detectibel then a single device, but I the theory woudl still hold. Considering the high levels of wireless signals that exist in 2070, the noise floor would be quite high. I could see ambiant signal levels being 5-10x what a typical PAN device generates. Without spread spectrum, it would be nearly impossible to transmit data, but since it is possible to recover signals below the noise floor with SS, the PAN system becomes possible.

Now if a corp were to build an electro-magnetic "clean room", inside a faraday cage, detecting PAN signals woudl be much easier. ANY RF energy (below a VERY low level caused by radioactive decay of the contents of the room, and thermal noise in the reciver, ect.) could be tagged as a potential threat. However, any electronics in the room would need to be individualy shielded, makeing the whole operation something you would do only in VERY high security areas. (Think the server room in Mission Impossible.)

and in the end the question is "is it worth it?".

is it worth putting radio scanners in every office so on?

how many false positives are we getting, ho much do maintaining and auditing the security setup cost us and so on.

problem is that most corps have a security budget of "plot". they have as much security that the GM feel like trowing at the player.

in theory one can make a office prefectly safe. the reason for it not being done more often is that people are supposed to work there.

most likely the radio scanners will detect a blip, indicate so to a guard watching the overlay and he will tell someone doing a rounds to go check whats going on.

sure, if this inside a high security area where there should be no signals at all at this time then it would be a high alert asap. but maybe they have had some lab techs or scientists come in late before as they have had one of those bright moment and want to get something done while the idea is fresh in their mind.

we should never forget the human element. to many false alerts will make the guards ignore the one true alert.

and this is the essence of the story about the kid yelling wolf wolf one to many times. when he needed the help, the people didnt come as they belived it was just another false alarm.

and a smart team can play on this. a hacker could break the system and set of false alarms. then maybe they will shut it down thinking its broken and have someone call repair in the morning. then the rest of the team breaks in

the human element of computer securitys, it will allways be its weakest link

I missed where you said it. You mentioned isolated wireless.

I agree with you that SR4 says wireless is the norm. We were arguing the merit in SR4's stance.

Agreed

I disagree with this, but we have both presented our arguments previously and in other threads.

I was being sarcastic.... we need a standard symbol for it ... I hate smilies.


The rfid challenge response, is a tech that is already in place around labs.

Eh actualy one could argue their safer now that i think about it simply because of chokepoint issues. However if its connected to the matrix its hackable

It seems to me that for a generally fixed facility that the sensor network could build up a fairly accurate time-dependent noise floor profile. This in part assumes that after regular hours that the printers/coffee makers/workstations go to sleep mode where their transmissions are limited to routine SYN requests. Coordinate that profile with the GPS feed from the security guards' Comms and you should be able to identify rogue devices eventually.

If nothing else, changes in the noise floor below alert threshold may cause the guards/drones to alter patrol routes. This seems like it wouldn't be a bad thing since the naturally occurring changes (solar flares, lightning, etc) in noise floor would result in a less predictable guard route.

1. Frequency hopping is less common today then SS. SS is used in 3G wireless phones (CDMA is one type of spread spectrum), WiFi, military radios, Over the air HDTV (I think, not sure on this one), and many others. I would expect that by 2070, SS would be the standard method of radio comunication. I originaly mentioned "frequency hopping" since it came before SS and was useful to introdouce the concept.

2. Well, you could do it, but the cost would be higher then what we use now. From what i have read of 2050-2070 buildings are made mostly from plastics and composite materials (plasticrette) with possible metal re-enforcement. It would not be DIFFICULT to insert a metal mesh into a wall to block wireless signals, and I can see many security-minded companies doing this. However given the wireless nature of things this could be an imediment, by blocking all traffic into or out of a given room or set of rooms you effectively isolate that area from all electronic comunication. In a VERY high security area this may be exactly what you want. (Some highly classified research labs today use similar strategys to foil TEMPEST attacks reportadly) However, the expense is not usualy justified.

A quick websearch turned up this article:
http://networks.silicon.com/lans/0,3902466...39121501,00.htm

I am most interested in the "selectible" nature of the material. It could be left off during the day, when workers are actively monitoring the system and need outside access, then the area could be isolated at night when no activity is required. Remember, to fully isolate the area all windows, doors, HVAC ducts and other openings need to be covered with RF shielding. The shielding can be as simple as a fine metal mesh inbedded in the window-glass, but it must be electricly grounded to the rest of the room for maximum effectiveness.

In response to your sugestion theat RF detectors would appear at major traffic points: It would seem to me that these would be the LEAST effective places to use them. High-security areas, absoloutely, but in lower security areas with high traffic the high level of EM noise in the area would make it much more difficult to isolate any individual signal. Plus, you have to filter out legitimate traffic somehow. You are righ that in lower rf enviroments the "Type I" / "Type II" tradeoff becomes easyer. You still have to make the tradeoff, but you can get fewer false negitives for the same number of false positives then you could in a high rf area.

I think so. But since those are wireless now, they can be hacked! So high-sec areas could have both.

P.S. Please excuse all spelling. I am lazy and ignorant.

I agree, with some reservations:
1. The source remains in range and active for "enough" time for the detector to accumulate enough samples that it becomes statisticly unlikely to be random.
2. The noise floor profile is fairly constant. That is the equipment in the room, it's locations, and activity level changes only rarely.

The "eventualy" you mentioned can be repersented by a high threshold on the number of successes needed by the scanner to isolate that there really is something there. This is even MORE true if the building is **edit: not** at least partialy RF isolated, eliminateing stray external sources for noise, leaving only the stray INTERNAL noise sources. (includeing radioactive decay of the building and it's contents, among a few dozen other factors)

Very true. Of course a clever runner could generate some EM noise in one area of the facility (below the alarm threshold) and redirect the guards there while the team went in somewhere else...


Fun Fact: The original spread spectrum technique (really more frequency hopping, but the basic idea was the same) was developed by actress Hedy Lamarr in 1940 while she was playing piano. She was listed as the co-inventer on the patent.

IANLebo77 but I know some of these.

1. SS is in use right now with 802.11 b & g. (Technically Direct Sequence Spread Spectrum DSSS) 802.11b also uses a compression code key sent during preamble while g uses orthogonal frequency division modulation (OFDM) which uses sub-frequencies. OFDM will rotate the sub-frequencies to compensate for interference but not for privacy.

A quick googling says that the military uses frequency hopping systems that may also include OFDM and spread spectrum. Read articles here and here.

2. Not as a general rule but it is out there. Remember that mass blocks signals so dense materials are natural shielding agents. Missile-proof buildings are virtually EM proof as-is.

The current goal is selective filtering; blocking wifi but letting cellphones work. In SR4 there is no difference so no selective filtering. Simple shielding isn't that difficult to achieve but can be costly. The military discusses it here.

something like lighting or solar flares would be detected across multiple sensors.... therefore you can rule out the spike ... If you have a spike on one sensor .. then you investigate.


Edit: instead of posting again.

If SSS is in use now with wireless networks ... then how come it is so easy to listen to these packets? If you sniff 5 billion packet even with wep 128, you can decipher the key and then decode the data. Is it because there is no algorithim for the variance to the frequecies used based on the id used?

This all relates back to detecting a signal.

DSSS is intended to reduce sensitivity to interference and provide redundancy rather than privacy. A DSSS requires less total power than a single channel broadcast to have the same effective clarity of reception. DSSS in 802.11b/g just means you can have a weaker transmitter and still be heard. The 802.11b/g signals have a range of 400m, or Signal 3. This is not "stealth."

WEP, the security of 802.11b, was flawed in 2 parts. 1) you can capture enough traffic (~15GB) and learn every encryption pad and 2) was restricted to 40bit and 104bit RC4. No.1 means you can simply learn all the codes and then decrypt the data and No.2 is equivalent to having a combination lock that only needs 1 or 2 numbers to open.

802.11g uses AES encryption (stronger than RC4), increased the number of pads by a factor of 16 million so it will take petabytes of data to capture the pads, and uses full 128bit keys. Still hackable but harder.

The SS being used in this case is not designed to evade detection. It is being used to help cope with interferience, and multiple access. Also, the "sequence" being used to spread the signal is public. Were this not the case one would first have to determine the sequence being used, (difficult, since the signal looks almost exactly like broad-spectrum white noise). The question of cracking WEP is not on point since that is a method of encrypting the data channel, not anything that changes how the actual radio trasmission takes place. These systems exist on diffrent levels, with WEP running on a SS platform. The WEP could be implimented over a diffrent set of hardware, say... wired ethernet, or FIDI, or whatever.

Take a look at Wikipedia's articles on spread-spectrum:
http://en.wikipedia.org/wiki/Spread_spectrum

Also, consider that this type of comunication could be conducted in psudo-random "burst" modes. This, combined with spread spectrum systems could confuse an scanner.

Oh, I agree, (and am also an engineer). I was just trying to explain that the rules as written in the book do a fair job of presenting a possible future. There was a lot of chatter about how it's impossible, that there will be too much radio traffic trying to share too small a spectrum, and that anyone with a smartlink would trigger every scanner in a 1km radius. I felt it was important to present an approach to explan how these rules COULD aproximate some level of reality.

So it could work on some levels .... but would it be something that would it be deployed on a massive scale? There are many things that could work, but should not be because of scalability problems. I never doubted that it could work ... just that there are some major drawbacks from implementing them.

excelent summary of my view allso

hell, maybe one should make it a sig

anyways; given that more and more stuff have their internal workings set i firmware rather then hardware, how hard would it be for say a hacker to mod a system so that it uses a nonstandard way of jumping from frequency to frequency?

or how about your comlink have a default channel for signing stuff up in the PAN while it inserts new frequenzy jumping sequences at semi-random intervals in the normal comunication?

same can be for active and passive mode vs hidden. on the first two it have the default channel open, but on the last one is ignores the traffic one it unless the user have initiated some sort of connection outwards.

on active is transmitts a burst ever so often saying its there.

on passive it will reply if something sends out a burst to request info from everything within range.

on hidden it will ignore all traffic that was not initiated from inside. kinda like a "stealth" firewall in a way. only way you can detect its there is when someone connects to the outside from inside it

ie, having the comlink in hidden mode, avoiding contact with the outside unless absolutly needed, and having gear that changes its frequenzy patterns more or less at random and you should have something thats more or less undetectable from background noice unless its basicly standing on top of the detector, right?

"Would" depends on the conditions. IF there was a crash due to some program that caused network gear to physically die (my understanding of Crash 2.0), causing a shortage of networking gear while some company had a stockpile of prosumer hardware they were about to roll-out worldwide, hardware that's been publicly tested for a while, hardware that would get you up and running .... Yeah.

You do what you gotta do to make payroll. Once it is up and running and people start enjoying the benefits it is hard to go back to wired.

Even without a crash, I can conceive of a mostly wireless network scenario. Lots of organizations rely on VPN for their WAN services and the difference between open internet and open air is not that much in terms of risk as you should assume any systems you don't control have already been compromised. If your ISP's routers have been compromised (and there've been enough router OS flaws to say that they probably were vulnerabe) your communication can be intercepted and decrypted.

It does introduce a vulnerability in the office however if you already let people VPN in from home it isn't a new vulnerability. Wireless is just easier for people to understand the vulnerability that already exists.

Sorry, out of my expertise. I'd say it is unlikely b/c the computational systems may not be there.

But I don't think it would do what you want. The frequency cycling has to use a sequence that is anticipated by the other users on the network. Current military radios are synched before an operation so that their clocks are on the same time and so they all have the same hop algorithm; I see no reason for that to change in the future.

Really not sure what you want here. I'm thinking that this functionality would be easier addressed by spending the 50Y to add skinlink to the device and just link it to your PAN that way.

Gee, I'd have never guessed from your posts.

Yeah, I did the same thing with the "troll sized bow" a few days back. Surprisingly, laws of physics say "yes, a bow with a pull 3x as heavy and draw twice as long as an english longbow has a max range of around 900m."

I'm still a little sketchy on the effectively infinite amount of bandwidth they bandy about. Seems like interference and noise pollution will create a cap. But if that's my quibble, pfah. I'll live.

The thing people seem to lose is the notion this is an approximation of reality. They get hung up on how it's not exact. It doesn't have to be. Heck, I run hydraulic simulations and those are chock full of estimates, educated guesses, and just plain hunches. If my model results in an acceptable mass balance and time-history error, my two benchmarks, then it is obviously good enough.

If SR4 results in an enjoyable game, the only real benchmark to a game, then it is obviously good enough.

think of it this way then. inside most items outside of rfids (those have only one task, to spew back a signal if it gets the right signal) there is a chip and a clock. the chip can be programed to follow a sequence of frequency hops based on a sequence of numbers given to it at some time or another while the clock can be synced by a outside source by same signal.

so when you say pair a gun with a comlink the comlink squirts over the first hop sequence and allso a time sync so that the internal clock of the gun is in sync with the comlinks. at some interval within these hops the comlink will send out a new hop sequence to take effect on the next hop, and allso send a new time sync.

most likely the chatter between gun and comlink is allso encrypted by public key encryption exchanged when paired and maybe replaced at some interval.

so to infiltrate this stuff you will have to break the encryption before they decide to change hop sequences, and pray to god that they dont change keys while your trying to relocate the sequence

this in combo with short range of the pan signal should make it hard to crack and maybe even detect as you cant realy use pattern recognition to look for the signals. just when it locks down a pattern it changes

so, a better bet is to go in via the comlink as its the heart and brain of the pan and is allso set up to talk to other comlinks and computers at a longer range.

this you do over the public linkup frequency. then when the handshake and exchange of keys between comlinks are in place they to start using most of the same stuff that a comlink and a pan item uses. replacing hop sequences at intervals and allso public keys at intervals.

a comlink in hidden mode however will not accept a linkup whatsoever. it will just drop any incoming traffic of that kind unless its on its known good list (similar to setting up a wifi hotspot with a list of mac addresses.

basicly it will look as if its not there unless it sends out a ping to look for nearby equipment or similar. basicly it have gone radio silence but keeps listening.

a comlink on passive on the other hand will accept or deny this traffic rather then dropping it. thereby signaling that its there if asked.

a comlink in active mode however will keep broadcasting its pressence to everyone that listens at regular intervals. kinda like getting up on a chair and shouting that your here and wants to talk to anyone that cares to listen.

the reason i guess for military using only set hop sequences is that if a radio is lost then it will only be usefull to the enemy as long as they use that sequence. still, you could allso just drop said radio of the list of radios to talk to or replace keys. but its an extra layer of security. and the only time you do not add extra layers of that is when it gets in the way of doing the job its supposed to do.

I can almost see the look of pure joy on the face of the hacker who DOES get inside a network of a team carrying wireless grenades.

hehe especialy for the team joker who stuffed two down his pants as a joke

if they're wireless grenades i'm sure that could be done wirelessly...ie no actual physical pin but a wirelessly controled timer/detonator

While in radio-silence mode, a runner should probably have all sensors (eyes, ears, biomonitors, etc), analytical cyber (e.g. spatial recognizers), cyberlimbs, weapons, and mapping (GPS+mapsoft) active on their personal PAN.

A really good mapsoft with image recognition software could update the map based on the input from cybereyes. Highly accurate maps, like one that includes security systems, should be good for a +2 bonus to stealth checks.

If someone hears a guard and has spatial recognizers +map/GPS, they might get a +1 bonus when hiding b/c they can keep things inbetween the guard and the runner.

Cyberlimbs should be on the PAN just so the character knows if they get damaged or shut down due to an electrical attack.

Once you break radio silence, sharing vague data (there's a guard over that way") between mutiple GPS-enabled sources can result in "grenade accurate" triangulation.
Having a medkit online is good since it means the best doc in the party can run it remotely. Biomonitors provide a boost to the med check.

After a fight, repairing cyber and devices should be easier if you have access to logs and internal diagnostics.

Either you maintain radio silence or you have your PAN active - pick one.

If everything is cybered, there is no need for a PAN as cyberware can be linked to each other.
If some parts are external, skinlink is the way of choice.
Judging from the capabilities of ware and mapsoft, there would be no need for an additional image recognition software.

Usually, DNI translates such defects into... pain.

Biological input error.

PAN = Personal Area Network.
A network of Skinlinked devices is still a PAN, it justed uses a physical medium.

I don't make assumptions about the gear being cybered or not. Plus, mapsofts would run on your Comm in most cases which is external.

I'm not sure about the image recognition. I imagine that most mapsoft supports simple editing by the user but to actually update itself...... I'm not sure.

Pain means "I'm damaged" but it doesn't say "I've lost 30% of structural integrity." Pain is also a distraction in combat; hence pain editors and damage compensators.

Gnarg.
In SR4, the TT of PAN refers to a wireless one in 99% of the times...

True... BTW, just checked - there is nothing that prohibits you to run the mapsoft directly on your cybereyes - internal Commlinks are cooler, though.

The latter is usually something to worry afterwards, which you already mentioned.

True, but the fine difference between 'pain' and 'too much pain' is what keeps people alive... making Pain Editor a very dangerous implant, while Damage compensators don't eliminate pain entirely.

The whole idea of DNI is to use ASIST to generate the most natural feeling and handling possible.

And as I keep saying over and over again, shadowrunners are not normal people and will constantly live in that *EDITED* <1%. */EDIT*

I'm not sure. My books aren't handy but I wouldn't think that a full mapsoft would run on your eyes. You could load a static map into the eye's memory, sure, but add search, filtering, editing, GPS triangulation, etc and I think you need a real comm for processing.

Not arguing that DNI makes something instinctively comprehensible but cyber is different from meat. If you don't patch up meat it will die and take the rest of you with it. Damage to cyberlimbs doesn't typically pose a direct threat to your health, nor does it cause the rest of your cyber to shut down.

Non-hydraulic cyber will likely continue to function in a degraded state indefinitely (This assumes that the muscles/motivators are some piezo-like material that flexes under power). Hydraulic cyber will "bleed out" and become nonfunctional when it runs out of juice.

Were I a runner, I'd have cyber limbs tuned to limit DNI pain to "unpleasant ouchie" and for anything more severe send an AR feed.

maybe it should be <1% ?

*Smacks forehead*

Yes, that would be correct.

Yes because we know all shadowrunners opperate completely out of the norm 100% of the time.

Come on now. All shadowrunners do not fall in that 1%. Just as all normal people do not fall in that 99% range. What thats saying is, 'hey when we refer to a PAN we're talking a wireless PAN because virtualy no one has a wired one'. The fiction in the book flatout proves your statement is not 100% correct as most of the time, from what we can see they're opperating with wireless PAN's not hard wired ones. Shadowrunners are not exceptions to everything. You know 99% of people in 2070 sleep in some form of bed. Does that mean since shadowrunners are not normal people that they don't sleep in any form of bed? 99% of people in 2070 do not deal with cash at all. Does that mean all shadowrunners will be buying all their eq and such with script? 99% of people in 2070 use the matrix. So does that mean shadowrunners don't use the matrix?