#sma comp6441

[puneethkpati]

Something Awesome: RFID

Here we have reached the pseudo official end to this project for my university course.  I just did my presentation in the lab the other day and boy oh boy... people liked it!!! Woo hoo, I did something interesting!

Don’t be so disappointed though, this by no means indicates the end of my pursuit to crack these infamous MIFARE Cards. Reflection: Looking over my project, here are some points I think are worth noting... 1. I tend to jump into commitments too quickly before doing my research. (Bought 45 different RFID cards) 2. Starting the project was the hardest for me physically. But once I found my stride there was no stopping me. 3. I get demotivated easily with failure. Need to work on my resilience and persistence. 4. I also tend to stop progressing when I realise my skills are lacking. Instead of diverting the focus in the project and working on something else in the project, I give up completely and do something else. These are definitely something I can work on. And definitely, get better with. Marks: HD would mean I’d finished my goal. Which I didn’t. But I do know how I could do it. The distinction would mean I’ve covered most of my goals.  Judging from my overall progress. I’d say I got myself a solid Distinction!!

[puneethkpati]

Something Awesome: RFID ARDUINO && MFRC522:

I’ve chosen to work with a micro-controller and an RFID read/write module.  Brief Overviews: Arduino is a handy little microcontroller used for a LOT of handy little projects. It’s great to work with sensors and IO modules as such. It really is a wonderfully powerful tool to have.  I write my code utilising the several libraries and modules on Github on a Proprietary Arduino IDE. This code is then uploaded onto the controller. All the code that the libraries were built on and I have extended is in C++... I’ve never coded in C++... MFRC522 is an RFID read/write module for Arduino, which can read or write to RFID cards. The cards supported really depend mainly on the code that is uploaded on to it from my understanding.   It works mainly on SPI technology and is a handy module for my handy little tool. 

[puneethkpati]

Something Awesome: RFID 

Encryption pt.3:

Sooooo.... yeah....

Crypto-1 was absolute garbage. It is to be expected when a company covers up a bad encryption algorithm with proprietary secrecy (Security through obscurity) Crypto-1 was so bad that the only challenge was reverse-engineering the actual encryption process. The rest was child’s play.

Upon the main weaknesses in Crypto-1, these are the ones that I read about! 1. Keys with their length’s of 48bits were Brute Forcible! (Someone got it done in 10Hrs) 2. Their Random Number generation process was predictable as f***  With the second weakness, most hackers found an easy way in... Considering blogging about this weakness too!

[puneethkpati]

Something Awesome: RFID

Encryption pt.2 Enter New Boss: Crypto-1

Remember how we thought there wouldn’t any scope for good enough Cryptographies to satisfy the versatile use of RFID tags and also offer complete security?

TURNS OUT WE WERE WRONG!

NXP Semiconductors had introduced their state of the art proprietary encryption algorithm! 

Crypto-1 changed the world of RFID and is perfect and will live on forever... I’m kidding... 

It’s Absolute trash...

[puneethkpati]

Something Awesome: RFID

We have covered, Memory Maps and this time I want to cover the individual type of blocks that cards have in their Memory:

Manufacturer’s Block 0: Easily the most crucial block on the entire card/chip. This stores the UID of each RFID card and also some of the Manufacturer’s Data. You cannot write to Block 0, only manufacturers can write to it. Or at least... That’s what they think.

Data Memory Blocks: Just regular blocks of 16B memory spaces where common data can be stored. Any block that isn’t Block 0 can be written or read when given access. 

The Read/Write Access is determined by Authentication and then Access rights.

Sector Trailer Memory Blocks: The Fourth/highest block of each Sector has the Key A & Key B values and the Access Rights. 

Authentication requires knowledge of the Keys and the Access rights depend on the Access bits in the Sector Trailer. Sector Trailers cannot be written to either... OR at least... that’s what they think.

[puneethkpati]

Something Awesome: RFID

Memory Maps

Every RFID Card has its own RFID proprietary Chip. This Chip in addition to the protocols for communications and UIDs have their own applications of memory maps to store the above.  The Cards that I dealt with, 

MIFARE Classic 1K/4K Chips,

Had a memory map that is very well structured out to make the cards useable for multiple purposes with Authentication.

The entire memory map can be explored by going through a hierarchical order.

Memory-Map is made of a collection of Sectors. These are just a collection of Blocks. Each Block is a collection of 16 Bytes.

MIFARE 1K:  16 Sectors x 4 Blocks x 16 Bytes

MIFARE 4K: (32 Sectors x 4 Blocks + 8 Sectors x 16 Blocks) x 16 Bytes

[puneethkpati]

Something Awesome: RFID

Cryptography on RFID:

RFID cards do not have a lot of memory on them. This explains why they are very useful and cheap to mass-produce and maintain. 

RFIDs technically need only carry the Unique ID number in their memory so the memory size really fits the technology.  However, People are mean! RFID use comes with the issues of maintaining confidentiality and authentication. Mean people shouldn’t able to figure out what I’m transmitting with the RFID tag. Mean people shouldn’t be able to replicate my RFID tag either. So to counter Mean people, We need some sort of security and cryptography on our RFID cards! BUT HEY! With such low memory sizes, how can we ever employ cryptography on these tiny devices? To employ RSA or AES the cost for mass-producing these bad boys would really be counter-intuitive to the use of RFID technology. While these could be employed for high impact confidentiality or authentication cards, they wouldn’t be sustainable for the biggest use of RFIDs.

What options do we have? I wonder...

[puneethkpati]

Something Awesome: RFID

Passive Read-only vs. Read-Write tags: 1. Read-only cards have a small non-volatile memory chip which has a UniqueID. 2. They only transmit the single UID which is implemented into the tag memory during manufacture and cannot be changed 3. Read-Write tags can be written to. 4. Usually, they have more space than READ-only tags to store both UID and some extra space for data. Either for logging or other uses.

Some research I did over the next week was encryption on cards and how Read-only and Read-Write cards handled it.

Also, researched the parts I needed for the project and ordered them on the weekend of Week 4.