Function to encrypt all data to be sent

[codebreaker176]

Hi,

Based on other discussions of encryption etc , I created a function to encrypt all the data, so that whenever I call the function It encrypts.

static   nrf_ecb_hal_data_t aes_struct;

static  uint8_t aes_data[16];

static   uint8_t aes_data_decrypted[16];

//Initializing arrays

memset (&aes_struct, 0, sizeof(aes_struct));

memset (aes_data, 0, sizeof(aes_data));

memset (aes_data_decrypted, 0, sizeof(aes_data_decrypted));

uint8_t* encrypt(uint8_t* p_data)
 {



        //Initializing key

        for (int i = 0; i < 16; i++)

        {
            aes_struct.key [i] = i;
        }

        //Initializing nouncence

        memset (aes_struct.cleartext, 0xaa, sizeof(aes_struct.cleartext)); //todo: use more random data and add counter

        //Creating chipertext

        sd_ecb_block_encrypt(&aes_struct); 

        //Encrypt -> XOR chipertext with p_data:



        for (int i = 0; i < length; i++)

        {  
            aes_data[i] = p_data [i] ^ aes_struct.ciphertext[i];
        }


        //decrypt -> XOR chipertext with Encrypted data:

        return (aes_data);

 }

 uint8_t* decrypt(uint8_t* p_data)
 {
    for (int i = 0; i < length; i++)

        {  
            aes_data[i] = p_data [i] ^ aes_struct.ciphertext[i];
        }


        //decrypt -> XOR chipertext with Encrypted data:

        return (aes_data);
 }

But I am not able to call the function, there's always an error

uint8_t vib_cnt_buf[5];
Convert_to_ascii(right,(&vib_cnt_buf[0]),8);
&vib_cnt_buf = encrypt(&vib_cnt_buf); //Encrypting this data

Can you please help

[ovrebekk]

Hi

Are you able to provide some more information on the error you are seeing?

Is it the Convert_to_ascii(..) function that returns the error?

Are you able to share the implementation of this function with me?

Best regards
Torbjørn

[codebreaker176]

Hi again,

Error : lvalue required as left operand of assignment

this error is at this lilne : &vib_cnt_buf = encrypt(&vib_cnt_buf);

I can't seem to get rid of it at all

the convert_to_ascii() function works well and converts the same variable to ascii values 

[ovrebekk]

Hi 

Sorry, I didn't notice that line earlier, it is clearly an illegal operation. 

You are taking a pointer to a pointer to the vib_cnt_buf array and trying to assign a pointer to it, this will not work. 

Instead I would suggest doing something like this:

uint8_t *encrypted_data_ptr = encrypt(&vib_cnt_buf);
memcpy(vib_cnt_buf, encrypted_data_ptr, length);

Best regards
Torbjørn

[codebreaker176]

Hello,

There is another error in the above mentioned code at line 9 : memset (&aes_struct, 0, sizeof(aes_struct));

Error : expected declaration specifiers or '...' before '&' token.

Thanks for the support

[codebreaker176]

Hello,

I am now using ECB mode of AES, there were lot of errors in previous one.

Now in the below code for ECB mode I have just one error : which I can't figure out the meaning,

static void nus_data_handler(ble_nus_evt_t * p_evt)
{   
    //some codes over here

    Convert_to_ascii(Battery_voltage,(&vib_cnt_buf[0]),4);
    uint32_t *encrypted_data_ptr = ctr_encrypt(&vib_cnt_buf);
    memcpy(vib_cnt_buf, encrypted_data_ptr, length);
    err_code = ble_nus_data_send(&m_nus, (&vib_cnt_buf), &thi, m_conn_handle);
}
/* 
.
.
.
.   
.
.
.
 */
static uint32_t crypt(uint8_t * buf)
{
    uint8_t  i;
    uint32_t err_code;

    if (!m_initialized)
    {
	    return NRF_ERROR_INVALID_STATE;
    }

    err_code = sd_ecb_block_encrypt(&m_ecb_data);
    if (NRF_SUCCESS != err_code)
    {
	    return err_code;
    }

    for (i=0; i < ECB_KEY_LEN; i++)
    {
	    buf[i] ^= m_ecb_data.ciphertext[i];
    }

    // Increment the counter.
    (*((uint32_t*) m_ecb_data.cleartext))++;

    return NRF_SUCCESS;
}


uint32_t ctr_encrypt(uint8_t * p_clear_text)
{
    return crypt(p_clear_text);
}


uint32_t ctr_decrypt(uint8_t * p_cipher_text)
{
    return crypt(p_cipher_text);
}

The following line has error (line 47 in above snippet) : 

uint32_t ctr_encrypt(uint8_t * p_clear_text)

{

}

Error : conflicting types for 'ctr_encrypt'

[ovrebekk]

Hi 

The following line is definitely not correct:

uint32_t *encrypted_data_ptr = ctr_encrypt(&vib_cnt_buf);

First off, when you use the name of an array without an index you are essentially getting the address of the first element in the array. 

Put another way, vib_cnt_buf and &vib_cnt_buf[0] is the same thing. 

Using &vib_cnt_buf will not work, as you are trying to get the address of the address of the first element in the array. 

Secondly, you are defining encrypted_data_ptr as a pointer, but ctr_encrypt returns a uint32_t, not a pointer to a uint32_t.

I would suggest starting by reading up on pointers in C either from online tutorials like this one, or from a book such as "The C Programming Language by Kernighan and Ritchie". 

Best regards
Torbjørn

[codebreaker176]

Hi,

Had not noticed, that was a very fatal error from my side sorry.

Thank you Torbjørn!

By calling these two functions together does it give us the original data back?

ctr_encrypt(&vib_cnt_buf[0]);
ctr_decrypt(&vib_cnt_buf[0]);

I get a different (probably encrypted) data

[ovrebekk]

Hi 

Have you tried to run the debugger to see exactly what data the different arrays contain when running the XOR operation in ctr_encrypt and ctr_decrypt?

If the result of the ctr_encrypt operation is XOR'ed with the same chiphertext then you should get the same data back. 

If this is not happening I suspect you are doing something wrong in the handling of the different arrays. 

Best regards
Torbjørn

[codebreaker176]

Hi Torbjørn,

I checked the code now, there is no change at all in the data , I recieve the same data, it does not get encrypted.

I am using nrf tool box app, And I am trying to send an encrypted data to it.

I followed the instructions as per the below link:

https://devzone.nordicsemi.com/nordic/nordic-blog/b/blog/posts/intro-to-application-level-security-using-the-ecb-

[Turbo J]

codebreaker176 said:

By calling these two functions together does it give us the original data back?

Nope. Consider where the data output of these functions goes.

I get a strong feeling that your C programming skills are simply not up to the task. There are too many beginner questions/mistakes.

Consider working through a good introductory book (or course) - this 40+ year old language is NOT beginner friendly at all.

Note: Proper crypto is hard even for advanced programmers. There are waaay too many pitfalls. You are almost always better off using some higher level library - that solves problems you don't even know about yet...

[ovrebekk]

Hi 

I have to agree with Jörg here. If you are new to C development, then making your own encryption algorithm might not be the best place to start....

The only advantage of cooking your own encryption is "security by obscurity", which isn't really security at all. 

Have you investigated the security enhancements added to Bluetooth 4.2, namely the LE Secure Connections (LESC) pairing scheme?

This was added as a response to the admittedly insecure pairing used in the original Bluetooth 4.0 implementation, and uses an industry standard Diffie Hellman key exchange algorithm to safely distribute encryption keys over a non secure channel. 

The paper referred to in Daniel's blog is discussing the older pairing scheme in 4.0, and does not cover the enhancements in 4.2. 

Unless you are able to implement a security scheme better than this then there really isn't any point adding another (less secure) layer on top. 

Best regards
Torbjørn