**5. Aspects of multisensory data acquisition and processing**

Data necessary for driving a car comprised of various sources:


Physical measurements can be direct and indirect. The direct measurements are performed by various sensors while indirect measurements estimate values from other measurements, events, linguistic variables. The same physical value can be measured by different ways each characterized by different qualities of:


The events and linguistic variables are characterized by:


Thus the same physical value can be measured by different sensors and estimated indirectly with a great variety in accuracy, confidence, availability. For example, the car speed can be estimated using the car wheels with high availability and low precision; or from the GPS system with low availability, precision and high latency; or from a radar with high precision; or queried from other traffic participants with low confidence etc.

The wide variety of sources must be integrated using plausibility checks and inference based on the reliability and availability of the sources.

*Autonomous Driving and Cybersecurity by Design DOI: http://dx.doi.org/10.5772/intechopen.112042*

Since the sources may contradict each other, the inference model must support conditional reasoning. This is necessary when the confidence measure of a measurement is conditional on some event or other measurements, such as in the case of computed values. At the same time driving has mission critical aspect. Therefore it should allow reasoning under contradiction when different events and measurements contradict each other since so that contradictions be resolved using information from other sources available.

The confidence measure may describe either or both kind of uncertainty:


In addition to uncertainty the confidence measure also needs to describe contradiction allowing combination of erroneous sources.

Furthermore, the inference process is a subject of real-time constraints. Therefore the choice of must consider:


The latencies imposed on the measurement process consideration of the time aspect, such as the time stamps and time intervals of the values, events and linguistic variables.

### **6. AD-velocity and position control**

The performance of WLAN communication of multiple antennas is an important aspect in this context, especially as a MIMO system, to improve the channel capacities [11]. However, it essentially concerns the transport level. We consider it as a given [12]. And on the other hand we focus on the MIMO concepts for the control of the driving behavior of an ADS via the information chain [9]. For the present ADS with MIMO (multiple input, multiple output) characteristics [7, 8], we define the multisensory input information in a simplified way as follows:


The following should be considered as output variables:


To control the driving behavior of the ADS, the RTT of the information chain plays a crucial role.

Suitable methods for controller synthesis are available according to (Ackermann). For digitization, the choice of sampling time is

$$T = \frac{RTT}{2} \tag{1}$$

or sampling angular frequency is

$$
\alpha\_T = \frac{2\pi}{T} \tag{2}
$$

where ω*<sup>T</sup>* , according to the Shannon theorem, is the largest angular frequency occurring in the information chain. However, the angular frequencies of the disturbance signals, the multisensory input variables, and the bandwidths of the controls must also be considered in this context. These considerations apply to both control variables, ADS velocities and the continuous determination of collision-free positions.

### **7. AD-driveability**

Initially, driveability refers to a vehicle's driving dynamics, particularly in terms of power, throttle response, engine, transmission, braking and steering control. It is an important aspect of the overall ride quality of a vehicle and has a significant impact on driver experience and customer satisfaction.

Good driveability means that the vehicle responds smoothly and predictably in all driving situations. Driveability is particularly important in modern vehicles with electronic controls, as it ensures precise and responsive control of the engine and other systems.

Driveability is of high importance for both comfort and safety. For example, in critical situations such as emergency braking or quick evasive maneuvers, good driveability can help the vehicle remain stable and the driver to maintain control.

Autonomous vehicles are not driven by human drivers. Therefore, the term driveability should be redefined as AD-driveability. This creates a basis for objectively evaluating different MaaS and TaaS concepts in terms of driving style and experience.

As far as comfort is concerned, passengers should not be impaired in their activities (working, reading, sleeping...) during the journey. For example, by braking too hard, accelerating too fast or driving in a jerky manner.

Good and safe driving behavior "AD-driveability" will become a competitive factor for autonomous vehicles, as the purchase decision will essentially depend on it. It is expected that the MaaS, TaaS concept, which reaches the destination faster with smooth driving comfort, will achieve a higher acceptance in the MaaS and TaaS service market.

The solutions outlined here, for the correlation of real-time and cybersecurity and adaptive real-time managers can make a decisive contribution to this.

### **8. Conclusion**

This paper has presented a comprehensive overview of various challenges and potential solutions related to autonomous driving and cybersecurity by design.

#### *Autonomous Driving and Cybersecurity by Design DOI: http://dx.doi.org/10.5772/intechopen.112042*

Ensuring real-time control, end-to-end cybersecurity, and driveability are critical aspects of developing successful autonomous driving systems. The proposed adaptive-real-time-manager (ARM) concept is a promising approach to addressing these challenges by continuously assessing and optimizing the real-time capability of autonomous driving systems while considering various influencing factors and selectively integrating cybersecurity mechanisms.

The integration of edge computing, parallel key renewal, and authentication, as well as the adaptation to future crypto requirements, are essential elements for ensuring the security and performance of autonomous vehicles in the connected world. The Cloud-Broker-Concept and simulation of the Adaptive Real-time Manager and the Cloud Broker further support these efforts by facilitating the integration into an ADS system and allowing for more effective testing and optimization.

Aspects of multisensory data acquisition and processing have also been explored, emphasizing the importance of integrating a variety of data sources and managing uncertainties and contradictions in the inference process. Speed and position control have been addressed as crucial aspects of autonomous driving, highlighting the significance of considering the round trip time of the information chain in controller synthesis.

Finally, the concept of driveability has been discussed in the context of autonomous vehicles, underlining its importance for passenger comfort, safety, and overall user experience. As the field of autonomous driving continues to evolve, the strategies and concepts presented in this paper serve as valuable building blocks for developing secure, efficient, and adaptable autonomous driving systems that meet the demands of an increasingly connected world. Future research and development efforts will undoubtedly reveal new challenges and opportunities for further enhancing the safety, performance, and acceptance of these innovative transportation solutions.
