Hola, estoy trabajando en un proyecto en Java Card que implica una gran cantidad de módulos de multiplicación. Logré implementar un módulo de multiplicación en esta plataforma usando el criptosistema RSA, pero parece funcionar para ciertos números.

public byte[] modMultiply(byte[] x, short xOffset, short xLength, byte[] y,
        short yOffset, short yLength, short tempOutoffset) {

    //copy x value to temporary rambuffer
    Util.arrayCopy(x, xOffset, tempBuffer, tempOutoffset, xLength);


    // copy the y value to match th size of rsa_object
    Util.arrayFillNonAtomic(eempromTempBuffer, (short)0, (byte) (Configuration.LENGTH_RSAOBJECT_MODULUS-1),(byte)0x00);
    Util.arrayCopy(y,yOffset,eempromTempBuffer,(short)(Configuration.LENGTH_RSAOBJECT_MODULUS - yLength),yLength);

    // x+y
    if (JBigInteger.add(x,xOffset,xLength, eempromTempBuffer,
            (short)0,Configuration.LENGTH_MODULUS)) ;
    if(this.isGreater(x, xOffset, xLength, tempBuffer,Configuration.TEMP_OFFSET_MODULUS, Configuration.LENGTH_MODULUS)>0)
    {
        JBigInteger.subtract(x,xOffset,xLength, tempBuffer,
                Configuration.TEMP_OFFSET_MODULUS, Configuration.LENGTH_MODULUS);
    }

    //(x+y)2
    mRsaCipherForSquaring.init(mRsaPublicKekForSquare, Cipher.MODE_ENCRYPT);

    mRsaCipherForSquaring.doFinal(x, xOffset, Configuration.LENGTH_RSAOBJECT_MODULUS, x,
            xOffset); // OK

    mRsaCipherForSquaring.doFinal(tempBuffer, tempOutoffset, Configuration.LENGTH_RSAOBJECT_MODULUS, tempBuffer, tempOutoffset); // OK


    if (JBigInteger.subtract(x, xOffset, Configuration.LENGTH_MODULUS, tempBuffer, tempOutoffset,
            Configuration.LENGTH_MODULUS)) {
        JBigInteger.add(x, xOffset, Configuration.LENGTH_MODULUS, tempBuffer,
                Configuration.TEMP_OFFSET_MODULUS, Configuration.LENGTH_MODULUS);
    } 

    mRsaCipherForSquaring.doFinal(eempromTempBuffer, yOffset, Configuration.LENGTH_RSAOBJECT_MODULUS, eempromTempBuffer, yOffset); //OK 


    if (JBigInteger.subtract(x, xOffset, Configuration.LENGTH_MODULUS, eempromTempBuffer, yOffset,
            Configuration.LENGTH_MODULUS)) {

        JBigInteger.add(x, xOffset, Configuration.LENGTH_MODULUS, tempBuffer,
                Configuration.TEMP_OFFSET_MODULUS, Configuration.LENGTH_MODULUS);

    }
    // ((x+y)^2 - x^2 -y^2)/2
    JBigInteger.modular_division_by_2(x, xOffset,Configuration. LENGTH_MODULUS, tempBuffer, Configuration.TEMP_OFFSET_MODULUS, Configuration.LENGTH_MODULUS);
    return x;
}


public static boolean add(byte[] x, short xOffset, short xLength, byte[] y,
        short yOffset, short yLength) {
    short digit_mask = 0xff;
    short digit_len = 0x08;
    short result = 0;
    short i = (short) (xLength + xOffset - 1);
    short j = (short) (yLength + yOffset - 1);

    for (; i >= xOffset; i--, j--) {
        result = (short) (result + (short) (x[i] & digit_mask) + (short) (y[j] & digit_mask));

        x[i] = (byte) (result & digit_mask);
        result = (short) ((result >> digit_len) & digit_mask);
    }
    while (result > 0 && i >= xOffset) {
        result = (short) (result + (short) (x[i] & digit_mask));
        x[i] = (byte) (result & digit_mask);
        result = (short) ((result >> digit_len) & digit_mask);
        i--;
    }

    return result != 0;
}
public static boolean subtract(byte[] x, short xOffset, short xLength, byte[] y,
        short yOffset, short yLength) {
    short digit_mask = 0xff;
    short i = (short) (xLength + xOffset - 1);
    short j = (short) (yLength + yOffset - 1);
    short carry = 0;
    short subtraction_result = 0;

    for (; i >= xOffset && j >= yOffset; i--, j--) {
        subtraction_result = (short) ((x[i] & digit_mask)
                - (y[j] & digit_mask) - carry);
        x[i] = (byte) (subtraction_result & digit_mask);
        carry = (short) (subtraction_result < 0 ? 1 : 0);
    }
    for (; i >= xOffset && carry > 0; i--) {
        if (x[i] != 0)
            carry = 0;
        x[i] -= 1;
    }

    return carry > 0;
}



 public short isGreater(byte[] x,short xOffset,short xLength,byte[] y ,short yOffset,short yLength)
    {
        if(xLength > yLength)
            return (short)1;
        if(xLength < yLength)
            return (short)(-1);
        short digit_mask = 0xff;
        short digit_len = 0x08;
        short result = 0;
        short i = (short) (xLength + xOffset - 1);
        short j = (short) (yLength + yOffset - 1);

        for (; i >= xOffset; i--, j--) {
            result = (short) (result + (short) (x[i] & digit_mask) - (short) (y[j] & digit_mask));
            if(result > 0)
                return (short)1;
            if(result < 0)
                return (short)-1;
        }
        return 0;
    }

El código funciona bien para un número pequeño pero falla en uno más grande