Wednesday, September 19, 2012

Automated Research and Some More Wireless Foo

One of the things that I sometimes wonder is whether or not Australia is lucky or unlucky in being so geographically isolated. Some industries (Airline and Motoring are probably the most recent and notable examples that come to mind) have not been provided with a level playing field when compared with the rest of the world. They've either been playing with one hand tied behind their back or others have. They haven't learnt to adapt. I think this is what happened with another industry that I recently had interactions with. Overseas, firms in this particular sector have advanced significantly over local competitors most particuarly in the clever use of technology. I've been thinking about this issue more though. Some of the concepts mentioned in my previous report (Convergence Effect)(such as 'Automated Research', have had significant resources allocated to them but it's becoming clearer to me just how far we can push it. Data mining, searching, statistical analysis, and report writing are all processes that can be significantly automated with sufficient knowledge and application of technology (I'll provide a more detailed framework on how to achieve this in the context of Security analysis in my report on, 'Cloud and Internet Security' (about 557+ pages/155K+ words now) but obviously it can be used in other contexts as well).

In my recent experiments I've obviously been experimenting with various different USB wireless adapters. One of them was the D-Link DWA-125 A3 (rt5370sta), another was the NetGear WNDA4100 (rt3573sta, you need to modify the driver to add support for the new chipset,, while another was Jaycar's YN8306 (all Ralink chipets funnily enough with decent Linux support, Clearly, after doing a lot of experimental work I've discovered a lot about antenna technology as well as the current state of wireless technology out there. One of the things I didn't realise is how drastic some of the differences could be between revisions of both adapter and chipset. My original purchase of the YN8306 seemed to be reasonable but the performance was nothing spectacular (Ralink 3070L). After discovering the USB port was somewhat loose (it was constantly losing grip of the cable) I exchanged it for another one but discovered something similar so this time I opened up the device and made some modifications (extra solder to bolster the strength of the USB port). While I was in there I noticed that the chipset had been upgraded to the Ralink 3072L. My thoughts regarding the Rev A were that it was good but didn't have the level of sensitivity that I thought it would. Rev B was a bit different though. Based on initial observations it may be at least 20% stronger with regards to sensitivity when compared to the Rev A and transfer rates were far less likely to fluctuate. A significantly better device overall.

With my recent incursions into antenna design I should probably extrapolate on what exactly I mean. I've torn apart anything and everything that I could possibly get my hands on and have even resorted to modifying television equipment in combination with broad range amplifiers to boost signal strength (I know it's not 'tuned' to the correct frequency but I just wanted to see what would happen.). What have I learnt? With enough imagination a lot of weird and wonderful things are possible. While hooking up my wireless adapter to the home television antenna (yes, the one on the roof) prooved to be an interesting exercise it only yielded slightly better results than having no antenna at all. A broad spectrum amplification device brought it up to slightly better than the originally included 5dBi antenna and forays into using antenna systems extracted from a dead router only yielded neglible benefits. A 16dBi antenna purchased during a local swap meet did result in a noticeable (though still small) gain though. Using a sieve (it's actually quite difficult to find suitable parabolic/spherical sized shapes in the house (unless you're willing to tear up potentially valuable goods such as headlamps). I had to go out and purchase a sieve for $2.80 at a local 'value store') only helped in certain configurations (I later found out an antenna from an old eMac was actually able to significantly boost the sensitivity of the DWA-125 A3 depending on it's polarity.) unless I switched to a more forgiving Gregorian Satellite Dish design (I used another smaller sieve and some elastic bands to replicate the design).

I even tried extremely crude Cantenna and Yagi style antenna designs (I'm waiting for an order to come in). The difference was that I made mine out of solf drink bottles with the bottom cut out wrapped with aluminium foil and also something I call 'Ruler-Yagi'. Basically, an antenna that was made out of a wooden ruler (boom), gold paperclips (it's suprising what you can find these days at Officeworks), blu-tack, and connected to a wireless adapter using alligator clips. There was a significant boost in signal (50%) when my 'Softenna' was directed properly and surprisingly there was an actual gain for 'Ruler-Yagi' but once again it was only minimal (10-20% signal sensitivity increase).

In the context of cyber-intelligence/operations it interests me to see what sort of antenna systems we could possibly pursue. There is a high likelihood that given the right materials high gain antennas could be embedded or inserted inside of suit cases or even other devices with connectivity being provided via cable (or hopefully) wireless. For example, a supposedly long life battery could be modified to provide a genuinely higher gain antenna provided space/cells were cleared from the internal cavity. Alignment could be facilitated by an internal lens/reflective camera system.

Sunday, September 16, 2012

Windows (and Android) Aren't (That) Inflexible...

If you've played around with Windows ISO's then you'd realise that with the advent of Windows Vista and 7 it is possible to convert a generic ISO to a so called 'Universal' addition through some ISO file modifications (and of course the correct 'Prodcut Key').

If you've been playing with Windows 8 and have not become accustomed to the interface there is some salvation.

For those of us who need or would like to extend the trial period for Windows 7.

If you haven't realised already the Kobo VOX is basically a low end tablet running Android. Moreover, there is an active community out there that is working on unlocking  the full capabilities of this device.

You can mount filesystems other than FAT32 on Android and have multi-partition support though I wouldn't say that it is 'clean' at this stage.

If you haven't heard already it is possible to run Debian alongside Android on your smartphone making the possibilities almost limitless.

Monday, September 10, 2012

Mobile (Wireless and Power) Foo

Recently, I have had to go slightly more mobile and in doing so I've been examining both portable wireless as well as power solutons. There has been some previous work with regards to powering a netbook using a series of AA batteries in series but I've been experimenting with basically using anything I have available to me. Seeing as though I'm limiting myself to what I have though it's been interesting. Some of the battery holders that are available aren't designed to work unless they are completely populated meaning some re-wiring work is required (alligator clips) if you don't have a perfect set of batteries to fulfil your requirement (yes, I'm aware of the recommendation that you should use a balanced set of batteries but I only require the supply for a maximum of 1 minute while I switch over the actual laptop battery. Another alternative is of course using a voltage regulator/inverter.

You may be asking why I'm not just using a power inverter or a car/air laptop adapter of some sort? Well, based on what's available at my local retailers the choice isn't optimal and doesn't come in at a reasonable price. Dirty power and modified sine waves just aren't my idea of an 'ideal solution'. Hence, my foray into more flexible forms of energy. A more refined product is of course available such as the following.

We're almost at the point where the price of solar panels, regulation, and storage have dropped (and cell efficiency has increased) such that it may be a worthwhile investment in a home (and mobile) environment. I've seen complete $200-300 options (panels, regulators, and storage pack) available at local electronics/automotive retailers of late.

Wireless technology has come a long way but it's still clear that more research is required. In some of my other work (my "Cloud and Internet Security" report is likely to exceed 600 pages now...) I've been scrutinising wireless security (and therefore long (at least 300 metres away between the source and the target with less than optimal conditions such as trees, fences, and other objects in between using both standard and inexpensive (less than $50) long range equipment. Look up Alfa wireless though there don't seem to be too many local options.) range and broad spectrum wireless options). There is still a substantial amount of Draft-N equipment (I've noticed some retailers have been clearing wireless equipment of late probably in anticipation of 802.11ac equipment. I've also noticed a trend towards merely modifying reference chips rather than seeking the most optimal performance and often manufacturers will swap chipset entirely between revisions if there are significant problems with the original release.) out there at the moment even though 802.11n was officially ratified several years ago and 802.11ac (pre-draft versions) equipment is actually on the market right now.
Based on random testing in my area it's clear that 2.4GHz wireless equipment is far more popular than 5GHz and dual band solutions are prohibitively expensive when compared with single band solutions. While MIMO has been around for a while multi-stream technology is still something we're figuring out. Three (two stream products seem to be mostly fine) stream products have only hit the market recently and research indicates that this may be due to the sheer complexity of such a setup and lack of sufficient CPU power (for a reasonable price) in a small package.

People seem to complain a lot about customer service/support these days. Based on an examination of local retailers there's very little to separate customer experiences. I have the funny feeling that if manufacturers/retailers just followed the letter of the law and provided reasonable products they would be considered the odd ones out? If you do some research you'll find that some practices border on illegality and depending on the firms in question there is a general lack of quality control and monitoring when outsourcing work.
Prior to some of my recent antenna experimentation I thought antenna design was reasonably simple. How naive... While there are many general designs out there a lot of them undergo extensive 'tuning' and it's not immediatelly clear why they are so different from 'average antenna' designs. For example, I recently opened up a USB wireless adapter, a DWA-125 A3 (removing RF shield on a DWA-125 A3 reduced signal strength by perhaps 5% but its perceived difference in performance was negligible to be honest). It was a strange design (from my perspective) because when you ran a continuity test through it's not immediately clear that there was a distinction between ground and signal. Compare it with an examination of a Sierra 306 USB 3G (Telstra Elite 21) modem where there is a clear and distinct one to one correlation between external antenna connector and on board antenna systems (I've read that that one is for transmission and another is for reception with sometimes another option for an auxillary antenna in embedded routers solutions).

Once you understand the concept of metamaterial antennas though it becomes clearer how they work. By altering materials you can (theoretically) increase the isolation between ground and signal providing for overall stronger signal quality and basically increase the virtual size of the antenna. Something which is not immediately obvious if you are new to miniature antenna design (I've seen some other work whch also questions this design but bypasses/improves on it by allowing for connection of higher gain conventional antenna solutions).

You need to understand the theory behind the antenna in order to be able to find optimal reception whether or you will never be able to extract the optimum solution without extensive trial and error. Recently, I was fiddling around with various antenna solutions (anything that was immediately available to me from pots, pans, cans, TV antennas, to COTS solutions. Most designs/solutions are too specialised though and have too narrow a performance band. The most general wide band solution that I came across was based on parabolas and cutlery holders in a directional/perpendicular arrangemnt/setup with back panel of an dissassembled eBook Reader which had a passive cooling solution on it in the form of aluminium sheeting.) seives and other parabola like shapes with non-standard shaped USB wireless adapters. Convential orientation with the antenna (and dish like object) pointing direct at the target was significantly better (from 55 to 75% signal quality) but non-optimal than when the horizontal reference plane of the receiver was below the target and therefore the azimuth between the target and the adapter was increased (75 to 100% signal quality). Admittedly, this had partly to do with the physical objects between the adapter and the target and the interaction (absorption, reflection, superposition, and so on) between them (high school and early graduate level physics is useful here). However, if you understood how the waveform radiates from the source than you would have understood why this was a more optimal setup.

In the context of pentration testing and network design this knowledge would also be useful in dealing with 'blind spots' and more covert access points. While there is software that can predict/simulate (within reason) wireless signal interaction with architecture and signal gain of different antenna designs in the real world this can be cumbersome. Its easier more useful if you know most of theory off the top of your head, guesttimate in the field, and then use software to confirm/more accurately determine your guesttimations.

Some serious beasts when it comes to routers out there at the moment. Power consumption similar to netbook (but with performance to match).

Typical connectors for reference.

Huawei 3G - CRC9
Internal Laptop - U.FL/I.PEX
NetGear/D-Link - RP-SMA
Linksys - RP-TNC