Help - Search - Members - Calendar
Full Version: Building the Mesh
Dumpshock Forums > Discussion > Shadowrun
Heath Robinson
Before we Begin
This fluff is probably completely incompatible with the canon Mesh Matrix information. This is no unintentional mistake. I enjoy looking at alternative views of a situation and preparing this has been an entertaining exercise in feverish writing at the very least.

Please abandon your preconceptions of the Matrix 2.0 and how wireless is "meant" to work. I'll try to explain my thought process along the way.


Building the Mesh

Meshes are all about deploying a bunch of cheap hardware that provides wireless coverage with routing code designed to adapt really quickly to emerging conditions. Conditions like "endpoint has moved" and "can't talk to this guy anymore, something's in the way" or even "massive dead zone". Throwing enough of the hardware into the environment gives a constant high throughput network that extends throughout the deployment area.

Walking the Talk

The success of a mesh is predicated entirely upon being able to pack enough devices into the environment that routing traffic becomes nearly free. Early implementations relied on power efficient omnidirectional aerials that could get picked up by everything in the environment.

The fact is that these aerials didn't have a choice about talking to everything at once. Any data that doesn't need to pass to you is essentially noise that you don't need to hear and it interferes with your ability to listen for the things that you care about. Early meshes had significant problems with scaling and density limits that prevented them from being widely deployed.

I am now Network Jesus

The inevitable saviour was an unlikely candidate, it was a relatively innocuous improvement in battery technology that merely increased battery capacity by a few tens of percent. A completely innocent chat between a military engineer and a civilian projects engineer that turned to new developments and how the military project was testing the new battery for powering a RADAR phased array sparked a new idea in the mesh network project.

A phased array uses complex electromagnetic emmission dynamics to direct the resultant radio waves in a chosen direction instead of spreading uniformally through the environment. Like a directed aerial it prevents the signal from straying from the target. Unlike a directed aerial, a phase array can reorient the directionality in milliseconds. The price is massively increased energy costs.

The app you love was an accident

Part of the problem with utilising directed point-to-point transmission is that you need to know where you need to aim the beam. You still have to talk in all directions or share information about where particular nodes are. A hybrid solution was most successful in terms of permitting high density networks. Initially devices will indicate their position when entering a section of the network by sending out a wide area broadcast to indicate their presence and then building a better picture using information from devices it talks to.

Phased arrays cannot communicate in all directions and it's considered a must for wireless to support mobile devices with a minimum of fuss. This made it necessary for every node to be able to tell where it's facing and where it's going. Cheap accelerometers and gyros were already being produced for use in drones and the code to make use of them was bog standard already. Attaching them to the prototype mesh components and incorporating some specialised code to optimise rebroadcast behaviour gave the mesh knowledge of where each device is currently located to a very good precision. This made sure the mesh could stay in communication despite the limitations in communication vectors.

Every development team makes toys and when one of them hooked up a 3D renderer to a bit of custom viewpoint calculation code for the mesh they inadvertantly realised an old idea.

Reality gets a makeover

Augmented Reality introduces computer drawn objects into your vision of the real world. It's pretty clever and has clear uses. By 2060 AR was in use in a certain highly specialised environments to enhance situational awareness in key personnel like surgeons. The necessity for the wearer to stay in contact with some computational resources or sensor packages prevented it being widely used in security or every day applications, but for stationary and non-violent applications it found some use.

The deployment of masses of wireless-enabled devices that all know positions and velocities to a reasonably accurate degree makes employing augmented reality simple. Augmented reality was not the original intent of the mesh, but it made sense to add it to the package as an optional extra. Nobody on the development team would have predicted how popular it would become.

Thankfully, marketing teams did.

The speedy deployment

Every prediction of the launch response of the new devices pegged it as a home and workplace product that would slowly integrate into the mainstream. In the office it would allow workers to share and work in whatever space they found themselves in as well as offering improvements to safety. Security would initially slowly pick up on the benefits that AR offered them in information awareness despite the proclaimed insecurity of wireless transmissions in security settings. In the home the main focus was to be on entertainment and effortless appliance integration without wires.

Nobody could have anticipated the Winternight attack, but they scrambled to capitalise on the destruction of the computing infrastructure of multiple major markets. Production went into overdrive and upper management bargained worldwide deployment and specification release into a place amongst the stars. NeoNET became a mega.

Over time, the meshes deployed in various countries have adapted to fit the guiding principles of their host nations, even as certain standards are laid down for interchange.

Wealth Disparity in an Automated Society

Everything talking to everything is a fundamental change in the scale of things. Your kettle can talk to your commlink, and your commlink can talk back. People with the money live like they're surrounded by servants every waking moment thanks to electronic devices negotiating and acting on their behalf.

Those who can't afford that kind of labour saving device live much like they did a century ago. Technology never quite reaches everywhere. In 2070 that "everywhere" includes places a few blocks away.

Cultural Influences

Every system reflects its designer in manners both large and small. This is an immutable fact of life in the computing business. If your organisation has three major departments then your program will come in three major packages. If they communicate badly, then your software will talk minimally and perform worse for it.

Every society has ended up building the basis of their mesh differently and their particular ethos is now built into the mesh in ways that are quite intractable. A short primer on the meshes of several major locations has been prepared to provide a grounding in how particular sets of ethoses will impact the design. These represent only an overview, more detail can be extrapolated by factoring in the culture of a particular location and the influences it might have on their mesh design.

UCAS

The UCAS mesh is based first and foremost around the Commlink as these devices have found their way into even the lowliest of hands. The mesh density is quite low because of a preponderance of devices capable of providing conenction functionality in a wide area and there is a self-perpetuating cycle that prevents the few attempts to provide near field communications.

The use of high throughput "carrier" nodes that service large areas have rendered the UCAS networks somewhat vulnerable to service interruptions from attacks and sparodic noise.

Many UCAS cities are sculpted by the authorities and the sculpting is typically delivered by a downloadable made available by a local government system. Links to it are frequently hosted by travel businesses. Businesses frequently sell replacement sculpting packages as well.

Technomancers in North America are respected as equal, but potentially dangerous, citizens and treated as such. Individual cities and counties are left to their own devices as to whether to implement Technomancer registration programs.

Japan

Japan has an extremely high density mesh combined with distributed computing assets and massive sensor deployment. These assets are held in possession of a government-funded private company.

Individuals get AR access for free and purchase AROs that operate as interfaces to programs running on some device somewhere as if they were physical objects. The migration protocols for these objects are complex and based upon assigned priveledges to use computational resources. The priveledge codes for these are sold by the corporation to interested ARO producers. Each code is permitted for a single ARO only.

There is a healthy trade in "shadow" AROs that are not issued codes by the owning company, but have been created with exploits that render them highly priveledged. These are hunted down by the authorities and deleted.

Due to the massive computational and sensor deployment, commlinks are considered unnecessary, bare AR glasses are common instead, and AROs with autonomous capabilities react to the environment as if they were real. Feedback clothing and Touchlink augmentations have become popular in a geek subculture of "immersivists".

The Japanese government has established a regulatory body that operates within the confines of the corporate-run system, recognising the "eminent necessity of policing the abuses of a public resource by the ill-intentioned few".

Scuplting priveledges are available to tenants of buildings and the lack of guidelines on style have created a fascinating but discordant outlook in many business areas. Akihabara in particular is a whirling site of clashing colour and extravagent design. Residential districts tend to lie untouched.

Technomancers in Japan are treated as a strange occurence that are prone to engaging in criminal behaviour. They are required to register with the authorities and are then watched closely. The corporate internal policy towards emerged entities is currently unknown.

AIs emerging in Japan tend to perceive the AR environment as natural and are typically real-world savvy.

United Kingdom

The United Kingdom still maintains much of its Matrix 1 infrastructure and the private sector has spawned a number of interface providers that allow mesh-enabled devices to bridge into the existing wired infrastructure.

A thin but growing layer of a widespread mesh is slowly emerging as mesh-enabled devices start to become available on the streets, but uptake is still slow. Meanwhile, commlinks are extremely popular with professionals and children and they tend to be used for communication and personal computing purposes.

Residential districts are growing sparodic hotspots of high density mesh as appliances get replaced by mesh-enabled equivalents and the occasional computer buff deploys dedicated hubs and more esoteric devices. Frequently the latter of these are imported from foreign countries.

AR is rarely a shared resource, instead being used by an individual and suited to that individuals tastes alone. Enterprising businesses offer reactive sculpting services that promise to integrate the AROs they produce seamlessly into the sculpting of the individual. These see varying success and penetration.

Technomancers are generally not identified as distinct from other mesh users. When a Technomancer is identified in a community they tend to be subtly ostracised. Instead, the VR-rape phenomenon attracts far more heightened emotion from much of the populace.

The government has implemented a Virtuakinetic registration regime at the behest of the EU and recently added VR-rape to the list of crimes that add your SIN to the sex offenders register.



I hope you enjoyed reading it. I would like to see comments on the direction, writing quality and technical aspects of this. Further developments would be much appreciated.
InfinityzeN
Tagged for reading and thought when I get home. A quick skim over really pinged my interest.
kzt
This is one of those posts that really good, but exposes some interesting issues with the whole underpinning of SR. So if Megas are all so powerful and not subject to any outside control, how did NeoNET not get "acquired" via a "hostile takeover"?

Bill Gates: "Oh, I didn't get rich by writing a lot of checks!"
KCKitsune
QUOTE (kzt @ Jan 8 2009, 11:14 PM) *
This is one of those posts that really good, but exposes some interesting issues with the whole underpinning of SR. So if Megas are all so powerful and not subject to any outside control, how did NeoNET not get "acquired" via a "hostile takeover"?


kzt... just do what I do when some bullshit like this comes up... mash the "I BELIEVE!!!!!!"™ button and move on. I mean honestly, Native Americans have a rich history, but they do not have numbers. They couldn't take over parts of the US, and the bullshit that nukes just sorta kinda NOT working... that's pretty lame. If nothing else the US government would have resorted to MOABs and Daisy Cutters. Missiles would have worked too. What about plain old guns? They didn't stop working. Yes the Spirits would have caused some pretty big problems... trouble is that they have to manifest to do so.



The worst thing that gets me mashing the "I BELIEVE!!!!!!!"™ button is the idea of extraterritoriality. The idea that governments would give up their power to corporations is stupid. Yes the corps can go: "Well I won't build stuff in your country! I'll take my money elsewhere!" works as long as the government in question doesn't respond by saying: "Well, OK, just don't expect to sell any of your products here in this country. We're banning your company from this country." and then giving tax breaks to their competitor.

Think about Microsoft and the EU. Why doesn't Bill Gates withhold Windows from the EU? Because no matter how good your product is, there is ALWAYS someone with the next best thing and he's waiting for you to fuck up. Linux lovers would cream their shorts thinking of Bill not selling Windows to the EU.
Heath Robinson
Alright, I need to craft an explanation of why Transys-Erika avoided being bought out. Time to flex my hands.


Returns, Investments and Methodology
Market analysis initially pegged the mesh and AR technologies as a slow line of profit that would produce a return of no more than 8% per year, with a pragmatic estimate placing the expected return at around 6.7% if the market accepted security assessments issued by the holding corp. For the investment it required that was not considered profitable enough to attract megacorp interest. Megacorps have enough budget to perform vast investments that give far better returns with lower risks. Megacorp budgets are actually more in demand than smaller corps on account of there being more good options.

In light of this, the fact that there wasn't any megacorp interest is understandable. Why invest in a risky new technologies that do not yet have any proven returns and aren't considered secure? Transys had some plans for rolling it out to enhance their own workplace efficiency and benefit from possessing the loyalties of the talent that initially produced it (who would be more likely to be extracted or abscond if the corp were bought out), but their primary reason for investing further in the technology was that it compared favourably to some of their other options.

Storm in a bottle
Crash 2.0 was a golden oppurtunity and a nightmare rolled into one. With workforces diminshed, lines of communication jammed closed and the collapse of significant portions of the infrastructure for multiple nations industrial espionage temporarily kicked into overdrive. 'Runners retired on the money made from the extractions and infrastructure damage missions offered. Every manager with a grudge polished off their "petty cash" budgets and tried to bring guns to the knife fight that is business.

Of course, the megas primarily targeted their major competitors and potential upstarts, using the paralysis of the Corporate Court to target assets that would get them Omega Ordered any other time. They didn't consider business moves possible, assuming that everybody else worth considering would be as isolated and paralysed as themselves. The Transys-Erika merger prompted them to reconsider.

Rewriting the Rules
Transys-Erika (used as an unofficial interim name) recognised that they possessed the standing investments necessary to reap dramatic profits by capitalising on widespread public mistrust in the existing corporate hegemony and the computational network that it produced. They scaled up the mesh networking component production runs and prepared a tailored marketing campaign touting the benefits of a ubiquitous wireless network.

It was not long before their opening gambits were decoded and understood. Most of the megas attempted outright buyout. Novtech, held in disgregard amongst the halls of power, offered a "merger" that was closer to a buyout but offered a few extra bells and whistles that would at least look like a polite fiction that the tripatriate were all held equally. As the best offer they were going to get, Transys-Erika's shareholders voted in favour of accepting the Novatech proposal narrowly.

With additional sources of capital, the production was scaled up further. Deployment was naturally hampered by a multitude of runs from a variety of factions, but the AR-enabled security infrastructure that was hastily deployed to all key locations within Transys-Erika-Novatech proved a small edge over intruders that prevented any major setback.

Overturning the judgement of the Corp
As events heated up on the ground, the "merger" was used to maneuver attempts to throw the tripatriate out of the Corporate Court and force readmission procedures. The grounds were that the combined entity was no longer officially considered the same entity as Novatech. Novatech heads, being more experienced with Corporate Court, understood the intent and offered a bargain that dangled the offer of releasing the mesh protcols and the creation of an independant CC-run body that would hold official control over the technologies. In exchange they would take back their seat on the Court and assign the first head of the combined entity.

One of the numerous small print terms officially renamed the meged corp to NeoNET, effective immediately and guaranteed that the name Novatech would continue to be used to refer to the older corporation in Court documents.
Krypter
It's unclear what this is supposed to accomplish. Are you talking about beamforming and direction-of-arrival? Mesh networks have a high overhead cost in terms of energy and processing, and their excess redundancy makes them unsuitable for all but the most specialized applications. You certainly couldn't run a global internet based on mesh networks. As for cultural variation, well, it's a good idea but your writing seems to mainly reflect existing stereotypes.

QUOTE
Cheap accelerometers and gyros were already being produced for use in drones and the code to make use of them was bog standard already. Attaching them to the prototype mesh components and incorporating some specialised code to optimise rebroadcast behaviour gave the mesh knowledge of where each device is currently located to a very good precision. This made sure the mesh could stay in communication despite the limitations in communication vectors.
Heath Robinson
Yeah, most of the country flavour will be stolen from a lot of current things but that's a given for RPGs. It's impossible to actually predict where things will go, and using the impressions people have about something will make it more accessible. Still, I'd like to know why you specifically mentioned stereotypes.


SR4 handwaves the power and processing requirements (Using Fusion power. IN A CAVE! WITH A BOX OF SCRAPS!). Things about the directionality in transmissions are meant to address potential noise issues in denser meshes and privacy concerns that people have used to criticise the move to a wireless Matrix set up.

Phased arrays appear to be one implementation of beamforming. This thing was witten over the course of about an hour using only knowledge I had in my head so it's going to be quite a bit less informed than someone who has decades of experience in the industry. The reason I'm posting this up here is to get some kind of comment on practicality and writing quality (which I know is really crappy, I was quite sleep deprived when I wrote it) as well as extra suggestions.

As for Mesh networks being for specialist applications only, I'm not seeing that in the major examples that I've found. I've not heard any specific limitations that might prevent them from being more widely deployed.
kzt
As far as I know nobody really knows how to make a large-scale mobile mesh work (large-scale = hundreds of thousands of nodes). When you add mobility to the mix the complexity rises really fast. Trying to connect from one mobile device roaming across self-organizing mesh networks to another device roaming across self-organizing mesh networks seems really hard. I can't see how you could connect them without the equivalent of dynamic DNS being continually updated by the mobile devices at some fixed site that is world accessible.
InfinityzeN
QUOTE (kzt @ Jan 11 2009, 10:36 PM) *
*Snip* I can't see how you could connect them without the equivalent of dynamic DNS being continually updated by the mobile devices at some fixed site that is world accessible.

Or several such sites that are land-lined (going to go with advanced fiber here) together.
KCKitsune
Well Microsoft has done something with Vehicle Wifi (ViFi... very original ohplease.gif ) It was in this month's issue of Discovery Magazine. I'd look that up if I were you.
Heath Robinson
Self-healing meshes already allow devices to drop out of the network, and be added to it on the fly. A device moving is, essentially, a device dropping out of the network and being reintegrated in another position. Keeping track of position and directional vectors is intended to provide for predictive hand overs that will make this whole thing smoother.

The routing information is a larger overhead than for a wired network, but it's assumed that by SR4 it is minimal compared to the operations that are performed as a matter of routine. You need to use chatter to spread new routing information, of course, and actively back propagate new routes for stalled connections. I can't see any major problem with this.
kzt
QUOTE (Heath Robinson @ Jan 12 2009, 10:48 PM) *
The routing information is a larger overhead than for a wired network, but it's assumed that by SR4 it is minimal compared to the operations that are performed as a matter of routine. You need to use chatter to spread new routing information, of course, and actively back propagate new routes for stalled connections. I can't see any major problem with this.

As long as your comlink can maintain state to every other device in the entire world it would work fine. Like a 50 billion AS BGP table....
JFixer
QUOTE (KCKitsune @ Jan 9 2009, 06:30 AM) *
kzt... just do what I do when some bullshit like this comes up... mash the "I BELIEVE!!!!!!"™ button and move on. I mean honestly, Native Americans have a rich history, but they do not have numbers. They couldn't take over parts of the US, and the bullshit that nukes just sorta kinda NOT working... that's pretty lame. If nothing else the US government would have resorted to MOABs and Daisy Cutters. Missiles would have worked too. What about plain old guns? They didn't stop working. Yes the Spirits would have caused some pretty big problems... trouble is that they have to manifest to do so.


Dude. This is old school SR lore. The Great Ghost Walk. They were /walking through hails of bullets/ and the bombs were launched... and vanished in mid-flight. Just vanished. The NatAm Shamans were operating on a level not seen since the Awakening, tapped into huge, yawning resevoirs of magic, as if generations of unbroken faith were suddenly rewarded with limitless amounts of power. There were only a double-handful of agents taking over the bases, and they were wading through trained military officers with nothing more than knives. Because they were moving at 4 Init passes and untouchable by bullets, while the military was utterly outmatched, and unprepared, for that kind of power.

It was a massacre, and by the time the power levels returned to normal, they already had all the territory and all the missles, and all the control they could possibly need... and they still had trained Shamans against a country with only mundane recourse and very little even in the way of Wares yet.
KCKitsune
QUOTE (JFixer @ Jan 13 2009, 01:05 AM) *
It was a massacre, and by the time the power levels returned to normal, they already had all the territory and all the missles, and all the control they could possibly need... and they still had trained Shamans against a country with only mundane recourse and very little even in the way of Wares yet.


And you are telling me that the US government would have just let these people take over all of the Nukes in the area? I don't think so... if nothing else they would have launched them and then self destruct them. Sure these guys might have been bad, but honestly... they were limited in scale. They might have gotten some nukes, but not enough to make THAT big of a difference. And what about AFTER? You think that when mana levels when back to normal that these few Indians could have HELD that area? Not really. This is just FASA handwavium bullshit.
Heath Robinson
QUOTE (kzt @ Jan 13 2009, 06:02 AM) *
As long as your comlink can maintain state to every other device in the entire world it would work fine. Like a 50 billion AS BGP table....

You make a valid point.

Library of congress. On a chip the size of your fingernail.

You could keep a scratch cache of routing information that you overwrite on a least recently used basis. You can poll nodes in the cache for additional routing information as you need it. Yeah, not particularly optimal, but mobile wireless nodes makes optimal routing nigh impossible.

You're probably also going to maintain reserve routing information. Overheads, I said they'd be higher
JeffSz
QUOTE (KCKitsune @ Jan 9 2009, 12:30 AM) *
Linux lovers would cream their shorts thinking of Bill not selling Windows to the EU.


*pauses while reading*

...

Yep.
This is a "lo-fi" version of our main content. To view the full version with more information, formatting and images, please click here.
Dumpshock Forums © 2001-2012