4.3 Our Novel Solutions
4.3.2 Novel Implementation of Digital Watermarking Technique
In Publication IV, we presented a unique security and authentication technique to securely transmit healthcare images in smart homes based on digital watermarking. The technique includes introducing digital watermarking [44–47] in which a special information or possibly another digital content is embedded into the healthcare images at the sender’s side. As described in Publication IV, our technique will be another unique way to adequately enhance the security of these healthcare images. The purpose of embedding a watermark into these healthcare images is to serve the purpose of access control, copyright protection, and authentication. Healthcare images are very sensitive and thus modification of these images is not allowed (nondestructive) [44, 48]. The method must also be reversible, i.e., the healthcare images must be accurately recovered back to their original state.
Wireless networks are very important in realizing smart homes and mobile health systems, but the current state in terms of security is a major concern. As illustrated in details in Publication IV, there are two main processes involved in our technique:
I. Embedding process: This process takes place before transmission over the wireless network. The smart home user embeds a special information or another digital image into a healthcare image and then transmits it to the recipient, which in our case may be the consultant.
II. Extraction process: This process takes place at the recipient’s side after transmission over the wireless network. At this stage, the watermarked healthcare image is verified to determine its authenticity and if it has been tampered with or not. If after the verification process, the watermarked healthcare image is confirmed to be OK, then the original image and watermark are extracted and verified. Otherwise, the image is simply discarded.
The efficiency of digital watermarking method in enhancing security has been confirmed by previous academic researchers [44–47]. We have proposed its novel implementation in Publication IV and we believe that it will significantly improve the security level for wirelessly transmitted healthcare images in smart homes and mobile heath systems.
As mentioned earlier, our technique has two parts: the first part is the RONI selection in which the Region of Non-Interest is separated from the Region of Interest (ROI). ROI is the sensitive region of the healthcare images and must not experience any change. After RONI has been selected, then the second part, which is the watermarking part, is carried out. In the second part, the watermark is
may be degraded. Thus, RONI was selected as a suitable place to embed this watermark [44, 49–52]. In Publication IV, we presented a freehand selection of RONI, since we feel that it guarantees the best results in our application, considering that different healthcare images would be transmitted and the ROI for each image differs.
Once RONI is selected, DWT watermarking technique is performed on the image, in which the image is decomposed into four bands: one low frequency sub-band (LL, Approximate sub-sub-band) and three high frequency sub-sub-bands (LH, Vertical sub-band; HL, Horizontal sub-band; and HH, Diagonal sub-band), where L is Low-pass filter and H is High-pass filter [46, 50]. In our novel technique, we embed the watermark into the LL sub-band of the decomposed images, because it guarantees the best results in our application in terms of both imperceptibility and robustness of the healthcare images and watermark.
In Publication IV, we experimented our novel method in Matlab and tested the effectiveness using four healthcare X-Ray images provided by Kuopio University Hospital and a watermark. All images were in JPEG format. The four healthcare X-Ray images are Left Hip, Chest, Pelvis, and Leg of sizes 440x554, 605x568, 622x543, and 447x543 pixels respectively and the embedded watermark was the University of Eastern Finland’s logo with 75x75 pixels of size. Figure 4 depicts the original images and watermark before embedment.
As presented earlier in our experiment in Publication IV, Table 2 compares the original images with the watermarked images as well as with the images after the extraction process is carried out, while Table 3 compares the original watermarks before the embedding process with the extracted watermarks.
In our method in Publication IV, the watermarks were embedded into the healthcare images in such a way that they remain highly imperceptible and cannot be noticed with the naked eye, and the images experience no change at all. The histograms also depict the same results. However, when comparing the original watermarks and the extracted watermarks, we noticed slight changes in the quality of the extracted watermarks after extraction from the watermarked images.
However, the aim of this experiment has been achieved, since the exacted images with no change in any pixels’ values were recovered.
Our results in Publication IV affirms that the technique is robust and reliable and it can be implemented in smart homes and mobile health systems for authentication and enhanced security of the healthcare images being transmitted via wireless networks to the recipient.
4.3.2 Novel Implementation of Digital Watermarking Technique for Wireless Transmission of Data
In Publication IV, we presented a unique security and authentication technique to securely transmit healthcare images in smart homes based on digital watermarking. The technique includes introducing digital watermarking [44–47] in which a special information or possibly another digital content is embedded into the healthcare images at the sender’s side. As described in Publication IV, our technique will be another unique way to adequately enhance the security of these healthcare images. The purpose of embedding a watermark into these healthcare images is to serve the purpose of access control, copyright protection, and authentication. Healthcare images are very sensitive and thus modification of these images is not allowed (nondestructive) [44, 48]. The method must also be reversible, i.e., the healthcare images must be accurately recovered back to their original state.
Wireless networks are very important in realizing smart homes and mobile health systems, but the current state in terms of security is a major concern. As illustrated in details in Publication IV, there are two main processes involved in our technique:
I. Embedding process: This process takes place before transmission over the wireless network. The smart home user embeds a special information or another digital image into a healthcare image and then transmits it to the recipient, which in our case may be the consultant.
II. Extraction process: This process takes place at the recipient’s side after transmission over the wireless network. At this stage, the watermarked healthcare image is verified to determine its authenticity and if it has been tampered with or not. If after the verification process, the watermarked healthcare image is confirmed to be OK, then the original image and watermark are extracted and verified. Otherwise, the image is simply discarded.
The efficiency of digital watermarking method in enhancing security has been confirmed by previous academic researchers [44–47]. We have proposed its novel implementation in Publication IV and we believe that it will significantly improve the security level for wirelessly transmitted healthcare images in smart homes and mobile heath systems.
As mentioned earlier, our technique has two parts: the first part is the RONI selection in which the Region of Non-Interest is separated from the Region of Interest (ROI). ROI is the sensitive region of the healthcare images and must not experience any change. After RONI has been selected, then the second part, which
may be degraded. Thus, RONI was selected as a suitable place to embed this watermark [44, 49–52]. In Publication IV, we presented a freehand selection of RONI, since we feel that it guarantees the best results in our application, considering that different healthcare images would be transmitted and the ROI for each image differs.
Once RONI is selected, DWT watermarking technique is performed on the image, in which the image is decomposed into four bands: one low frequency sub-band (LL, Approximate sub-sub-band) and three high frequency sub-sub-bands (LH, Vertical sub-band; HL, Horizontal sub-band; and HH, Diagonal sub-band), where L is Low-pass filter and H is High-pass filter [46, 50]. In our novel technique, we embed the watermark into the LL sub-band of the decomposed images, because it guarantees the best results in our application in terms of both imperceptibility and robustness of the healthcare images and watermark.
In Publication IV, we experimented our novel method in Matlab and tested the effectiveness using four healthcare X-Ray images provided by Kuopio University Hospital and a watermark. All images were in JPEG format. The four healthcare X-Ray images are Left Hip, Chest, Pelvis, and Leg of sizes 440x554, 605x568, 622x543, and 447x543 pixels respectively and the embedded watermark was the University of Eastern Finland’s logo with 75x75 pixels of size. Figure 4 depicts the original images and watermark before embedment.
As presented earlier in our experiment in Publication IV, Table 2 compares the original images with the watermarked images as well as with the images after the extraction process is carried out, while Table 3 compares the original watermarks before the embedding process with the extracted watermarks.
In our method in Publication IV, the watermarks were embedded into the healthcare images in such a way that they remain highly imperceptible and cannot be noticed with the naked eye, and the images experience no change at all. The histograms also depict the same results. However, when comparing the original watermarks and the extracted watermarks, we noticed slight changes in the quality of the extracted watermarks after extraction from the watermarked images.
However, the aim of this experiment has been achieved, since the exacted images with no change in any pixels’ values were recovered.
Our results in Publication IV affirms that the technique is robust and reliable and it can be implemented in smart homes and mobile health systems for authentication and enhanced security of the healthcare images being transmitted via wireless networks to the recipient.
Table 2. Comparing Original, Watermarked, and Extracted Images.
Image
Name: Original Image: Watermarked Image: Image After Extraction:
Left Hip
Chest
Pelvis
Leg
Table 2. Comparing Original, Watermarked, and Extracted Images.
Image
Name: Original Image: Watermarked Image: Image After Extraction:
Left Hip
Chest
Pelvis
Leg
Table 3. Comparing Original and Extracted Watermarks.
Experiment number: Image: Embedded Original
Watermark: Extracted Watermark:
1 Left Hip
2 Chest
3 Pelvis
4 Leg
5 CONCLUSION AND FUTURE WORK
Security is a very important issue in mart home environments due to the sensitive nature of private and confidential data being transmitted via wireless communication links. The wireless technologies being used in the implementation of smart homes have security issues that could have severe security implications if they are not carefully taken into account. Therefore, identification of these security issues is crucial to taking the appropriate steps towards mitigating them and enhancing the security of the collected data within these homes.
This thesis presents our studies on how to enhance the security of transmitted data via wireless interfaces in smart home environments. All our practical attacks and proposed techniques were experimented and practically demonstrated to show that truly these threats are real and how efficiently our proposed techniques will work in securing transmission of data via these interfaces in smart homes.
Firstly, we provided an investigation into the possible security issues in Smart Home Systems. In addition, we analyzed smart environments with an emphasis on the security challenges of the wireless network interfaces being utilized in these systems and we proposed possible countermeasures to mitigate these threats. We also applied threat modeling process to our SEAL system to identify the assets and threats to the system and we examined how the SEAL system can be designed in a more secure way that will guarantee a maximum protection of data transmitted across the system.
Secondly, we proposed and practically demonstrated in our laboratory environment three (3) attack scenarios against ZigBee network, which is commonly utilized for data transmission in smart homes. These attack scenarios are based on utilizing several vulnerabilities found from the main security components of ZigBee technology. We demonstrated with experimental figures that attack against ZigBee-enabled devices become practical by using our three attack scenarios. In addition, we proposed novel countermeasures related to the integration of time stamping mechanism into the encryption process of ZigBee and the use of intrusion detection and prevention system in the network, we believe our countermeasures will render these attacks impossible if the beacon frame process is continuously monitored and time stamped used for sent and received messages.
Thirdly, we proposed a novel method, which will strengthen the Bluetooth pairing process and thwart the MITM attacks by employing Steganography, and we demonstrated experimentally the efficiency of this technique using mobile phones.
It is obvious that the security of Bluetooth pairing process is not adequately addressed by this current cryptographic method, as previous researches have
Table 3. Comparing Original and Extracted Watermarks.
Experiment number: Image: Embedded Original
Watermark: Extracted Watermark:
1 Left Hip
2 Chest
3 Pelvis
4 Leg
5 CONCLUSION AND FUTURE WORK
Security is a very important issue in mart home environments due to the sensitive nature of private and confidential data being transmitted via wireless communication links. The wireless technologies being used in the implementation of smart homes have security issues that could have severe security implications if they are not carefully taken into account. Therefore, identification of these security issues is crucial to taking the appropriate steps towards mitigating them and enhancing the security of the collected data within these homes.
This thesis presents our studies on how to enhance the security of transmitted data via wireless interfaces in smart home environments. All our practical attacks and proposed techniques were experimented and practically demonstrated to show that truly these threats are real and how efficiently our proposed techniques will work in securing transmission of data via these interfaces in smart homes.
Firstly, we provided an investigation into the possible security issues in Smart Home Systems. In addition, we analyzed smart environments with an emphasis on the security challenges of the wireless network interfaces being utilized in these systems and we proposed possible countermeasures to mitigate these threats. We also applied threat modeling process to our SEAL system to identify the assets and threats to the system and we examined how the SEAL system can be designed in a more secure way that will guarantee a maximum protection of data transmitted across the system.
Secondly, we proposed and practically demonstrated in our laboratory environment three (3) attack scenarios against ZigBee network, which is commonly utilized for data transmission in smart homes. These attack scenarios are based on utilizing several vulnerabilities found from the main security components of ZigBee technology. We demonstrated with experimental figures that attack against ZigBee-enabled devices become practical by using our three attack scenarios. In addition, we proposed novel countermeasures related to the integration of time stamping mechanism into the encryption process of ZigBee and the use of intrusion detection and prevention system in the network, we believe our countermeasures will render these attacks impossible if the beacon frame process is continuously monitored and time stamped used for sent and received messages.
Thirdly, we proposed a novel method, which will strengthen the Bluetooth pairing process and thwart the MITM attacks by employing Steganography, and we demonstrated experimentally the efficiency of this technique using mobile phones.
It is obvious that the security of Bluetooth pairing process is not adequately
pairing process during data transfer will be robust against MITM attacks.
Steganography hides the existence of this process by embedding the keys inside a cover image before transmission to the recipient. The whole key exchange process is unknown to the attacker, because the attacker in this case will not even realise the images contain hidden data, only the recipient will be aware of the content; this is a major uniqueness of our technique. In our novel method, only the key will be sent to the receiver at the first phase and the receiver will reply back to the sender with his key. After both the sender and the receiver sent stego image, which has the key embedded, a shared key will be generated, which is in half of the sender's key and half of the receiver's key. In the second stage, the shared key will be verified by both sides. A message will be created at the final stage and integrated into the stego image. The stego image will be extracted by using the shared key in order to view the message and exchange it to check the originality of the hidden message.
Finally, we proposed a novel approach to improve the security and authentication of healthcare images transmitted via wireless network based on digital watermarking technique in which a special digital image is embedded into the RONI sections of the healthcare images before transmission over wireless networks to the receiver. Then at the receiver’s side, the embedded special digital image is extracted from the healthcare image and verified to confirm its authenticity. We demonstrated with experimental figures the effectiveness and robustness of this technique by implementing DWT algorithm to successfully embed a watermark into the RONI section of some healthcare images. The watermarks were embedded in such a way that the ROI sections of the images were not affected and the integrity of the images was protected. Our results in Publication IV show clearly that this technique is very robust and efficient in providing authentication and enhancing the security of healthcare images and it can be implemented for wireless communication in smart home and mobile health systems.
Security of smart homes depends greatly on the security of the wireless interfaces implemented for data transfer in these homes. Currently there are several challenges in smart homes and mobile health systems and these challenges are currently in the news as several attacks on smart homes were reported in recent years. In our study, we have analyzed in details the security issues currently faced in smart homes and we have practically demonstrated the reality of some attacks against wireless interfaces used in these homes. We have also proposed novel methods to enhance the security of this wireless interfaces if implemented, and we believe that our results presented in this thesis will be a very useful tool for security researchers to further develop other unique approaches for securing smart environments.
Smart homes are likely to be applicable not only in healthcare, but in various
The problems we want to investigate in our future research work are concerned with the following issues:
1. The adoption of smart homes is likely to increase, not only for healthcare purposes, but in every sector that affect our daily living. Thus, new attacks against them are likely to be found. We want to further investigate more security
1. The adoption of smart homes is likely to increase, not only for healthcare purposes, but in every sector that affect our daily living. Thus, new attacks against them are likely to be found. We want to further investigate more security